Illinois Institute
of Technology
Libraries
Digitized by the Internet Arciiive
in 2010 witii funding from
CARL!: Consortium of Academic and Research Libraries in Illinois
http://www.archive.org/details/armourengineer04armo
THE
ARMOUR
ENGINEER
THE SEMI-ANNUAL TECHNICAL PUBLICATION
OF THE STUDENT BODY OF
ARMOUR INSTITUTE OF TECHNOLOGY
CHICAGO, ILLINOIS
VOLUME IV.
NUMBER 1
JANUARY, 1912
ILLINOIS INSTITUTE OF TECHNOLOGY
PAUL V. GALV!N LIBRARY
35 WEST 33RD STREET
CHICAGO, IL 60616
Copyright, 1912
BY
M. A PEISER
THE ARMOUR ENGINEER
VOLUME IV. NUMBER I.
JANUARY, 1912
THE FLUX OF LIGHT METHOD OF PERFORMING
ILLUMINATION CALCULATIONS.
BY WM. E. BARROWS, JR., E. E.*
The problem which the illuminating engineer is most often
called upon to solve is the determination of the number and loca-
tion of lamps required to satisfactorily illuminate a certain room.
Moreover, architects, builders, men in the other engineering pro-
fessions and in other walks of life as well, are often required to
decide upon a lighting system of greater or less importance. In
the solution of problems of this nature no definite rules can be
formulated to apply to all or even to a few of the many classes
of service for which artificial light is used. The situation is sim-
ilar to that of the tailor, who can tell one approximately how
many yards of cloth will make a suit of clothes, but who wants to
see his customer, take his measure, ascertain his preference as to
color and quality of the goods and the style of suit, before put-
ting the shears to the web. The success of the engineer depends
upon his ability to give satisfaction, and this involves, first, a
knowledge of the means of accomplishing his task, and second, a
study of the case and a knowledge of the application of the means
to accomplish the end desired.
The problems of artificial lighting are by no means simple.
They involve a study of the physical, the physiological, and the
psychological characteristics of the individual, the decorative and
aesthetic properties of the surroundings, and the business and
economic conditions of the time and place. The proper treatment
of these different branches would require an entire volume and
is far beyond the scope of this article. The particular phase of
the subject which the writer wishes to discuss is known as the
flux of light method of solving illumination problems. It should
be understood that such calculations must necessarily be of a gen-
eral nature and that a study and knowledge of the surroundings
and conditions will be necessary iii order to apply the results of
such calculations to practical problem, with engineering signifi-
cance.
It is obvious that in order to determine the number of lamps
required for an installation it is essential to know : first, the
*,Vssistant Professor of Electric-al Euninwriug-, .\riuum- Institute of Tocliuolo.iry.
THE ARMOUR ENGINEER
[Vol. 4, No. 1
amount of light available from each unit of the system chosen,
and second, the amount of light required for the particular class
of service for which the illuminants are to be used. In regard
to the amount of light from each luminous source we will refer
to the results of tests by the photometrician. These data are
usually in the form of polar curves plotted to polar coordinates
and show graphically the distribution of light in one plane (usu-
ally the vertical plane) around the source of light as a center.
These results merelV indicate the value of the candlepower in
definite directions around the source and have no significance as
a representation of the quantity of light. An interesting com-
90'
60'
¥?^
i^
\ viScc
___V.-^^\ y\ /
^
<^^A^^Z^
X"""^ /\ y^ \
"^v ^^I
_— ^^^^^-t"""^ \y^
/^ /
■^\ J\
7S'
60'
AS'
30° iS' O IS' 30'
The Armour Engineer.
Fig. 1. Theoretical Distributions of Light in a Vertical Plane Giving tlie Same
Amount of Light Flux, or the Same Mean Spherical Candlepower.
parison indicating the misleading conceptions likely to arise in the
study of polar diagrams is illustrated by the polar curves of Fig-
1 . The four curves a, b, c, and d represent theoretical distribu-
tions of light in a vertical plane. The maximum values in these
four cases are approximately in the ratios of 15-19-50-60.
However, if these curves represented the distribution of light
from luminous sources in a vertical plane the mean spherical can-
dlepower or the total flux of light would be the same for each.
The fundamental theory of this section of the subject is
based on the study of spherical surfaces, in junction with which
it becomes necessary to determine either the mean spherical in-
Jan., 1912J BARROWS: ILLUMINATION CALCULATIONS 5
tensity in candlepower or the zonal or total flux of light in lumens.
In these spherical calculations it is assumed that the luminous
intensity is equal in azimuth and varies only in one plane passing
through the source, which, in the following discussion, will be
a vertical plane.
If we assume the source of light to be surrounded by a
sphere of radius r with the source as the center, and further con-
sider this sphere divided into a number of zones in such a manner
that the illumination of similar parts of each zone is uniform, the
total flux of light embraced by a zone will be equal to the product
of the average intensity and the area of the zone. From a sum-
mation of these products for each zone, the total value of light
Fig.
777* Arntoui' Engineer.
Relation of Areas of Zones Shown Graphically.
flux emitted by the source may be obtained and this divided by
the area of the sphere (47rr-) will give the mean spherical candle-
power.
The light flux in the lower hemisphere will be the sum of
the products of the zones and their intensities and this sum
divided by the hemispherical area i2-nr'^^ will give the mean
lower hemispherical candlepower.
The area of a zone subtended by an angle embracing the first
fifteen degrees (15°) below the horizontal is 7.66 times as great
as the area of the zone extending fifteen degrees from the vertical.
With the same intensity in each zone, the total flux of light em-
braced by the former zone will be 7.66 times that passing through
the latter. If now the source of light is of uniform intensity in
THE ARMOUR ENGINEER [Vol. 4, No. 1
all directions, and by means of a reflector half of the light from
the zone subtended l3\' the first fifteen degrees below the horizontal
be redirected downward through the zone extending fifteen de-
grees from the vertical, the intensity in the latter zone will be
increased to 4.83 times its former intensity.
These results are shown graphically in Fig. 2 (2)* where
Case I represents the normal condition, and the shaded parts the
relative amounts of light in the two zones.
Case II shows graphically the relative amounts and intensi-
ties of light in the same zones obtained by use of the reflector.
Thus the quantity of light depends not only upon the intensities
in the dififerent directions, but upon the areas of the zones which
the various intensities illuminate.
The area of a zone defined by the angles (/ and ada is
2tt cos ada
and the quantit}- which it receives will be
Fi= I 2Tr I COS, ada
Oi and a., being the angles with the vertical which locate the
meridians determining the zone. The mean intensity for the zone
is equal to the total quantity of light flux divided by the area.
If the spherical surface is divided into n zones subtended
by equal angles, then the total amount of light from the lamp
will be
F = Fi+F2+ • ' • +Fn= I 2vli cos ada-\- I itrl^cosada
+Fn= I 2 tt/i COS c<ia+ I
t/o t/o+'^
n
■ I 27r/n_i cos ada-{- I 27r/n
+ •••• + I 27r/n_i cos ada-\- I 27r/n COS ada
n "^ n
If the intensitv is uniform in all directions we have
=r
27r/ cos ada
*The small numerals in parenthesis inilic-ate references in the liibliofrraphy at
the end of this article.
Jan., lyiil NARROWS; ILLUMINATION CALCULATIONS 7
and the mean si^herical intensity ,
/ cos ada
I 2irl COS ada I
Jo Jo
47r 2
Unfortunately, the law according to which the intensity varies
is too complex to allow the integration to be directly effected.
Hence it becomes necessary to resort to methods involving ap-
proximations.
It can be shown by spherical trigonometry that the areas
of the zones of a sphere are to each other as their altitudes. Thus
the luminous flux in any zone of an imaginary sphere surrounding
a source of light is proportional to
27r/(cos a,— cos Qn),
where a, — (7. is the angle subtending the zone of reference, a,
and a, being angles measured from the vertical, and / is the aver-
age intensity of illumination in that zone.
This equation forms the basis of the graphical methods oi
Rousseau, Kennelly, Macbeth, Wohlauer and others for obtain-
ing the mean spherical candlepower and the light flux in lumens
for a source having its distribution of light equal in azimuth.
These graphical solutions of luminous flux calculations were de-
scribed in a lecture by the writer given during the graduate course
in illuminating engineering at the Johns Hopkins University in
November, 1910, and published in Volume Two (page 625) of
the bound volumes of those lectures, to which the reader is re-
ferred for greater details of these methods. Wohlauer's diagram
with practical modifications by ^Macbeth and the writer will alone
be considered in this article.
By the "fluxolite" diagram of Mr. Wohlauer the value of the
flux of light in lumens may be obtained by simply adding a num-
ber of linear dimensions drawn to scale and multiplying the sum
by some constant. The value of this constant depends upon the
number of angular subdivisions of the spherical area.
It can be shown geometrically that the altitude and hence the
area of a zone is proportional to the sine of the angle, measured
from the vertical axis, which bisects the zone. Hence if the
imaginary spherical area be divided into n numbers of equiangular
zones assuming the midzone intensity to be the average for the
zone, then the flux in any zone will be
F = KI sin B,
THE ARMOUR ENGINEER
[Vol. 4, No. 1
vvliere / is the average intensity of the zone, 6 the bisecting angle
nieasnred from the vertical axis and K a constant the value of
which depends upon the number of zonal subdivisions.
Referring to Fig. 3, and representing the flux in successive
zones from the nadir by Fj, F^, etc., the average intensities by
/j, /o, etc., and the midzone angles by 6^, dn, etc., we have
/SCfSS^ /so' /3S'
15" 30"
The Armour Engineer.
rig.- 3. Wohlauer "Fluxolite" Diagram.
F^ = KI^ sin e^ = i^Lab,
F. = Kh sin Q. = KL^^,
F3 = Kl^ sin ^3 = i^Lef,
F, + F, + . . .F. = K (Lab + L,. + Ui +
etc.).
Thus the flux in any zone is equal to the horizontal projection of
its midzone intensity multiplied by the constant, and the total flux
in lumens is equal to the sum of the several projections multiplied
by the constant.
The mean hemispherical candlepower may be obtained by
dividing the value of the flux in that hemi-phere by 2ir, and the
Jan., 1912J BARROWS: ILLUMINATION CALCULATIONS 9
mean spherical candlepower may be determined by dividing the
vahie of the total flux by Att. The values of K for various
angular subdivisions are given in the following table :
Angle embrace of zone. . 5 10 15 20 25 30
Value of i^ 0.548 1.098 1.64 2.18 2.72 3.25
In the example just cited K is equal to 1.64. 9 being equal to 15°.
The polar diagram is constructed with vertical lines spaced
The Armaur CnginetK
Fig. 4. Macbeth's Flux Scale.
equal to the polar scale to facilitate the evaluation of the projec-
tions of the various midzone intensities.
By referring to the values of the constants given above we
will see that for zones subtended by ten degree angles the value
of K is 1.098. If now the polar curve, Fig. 3, was plotted on
polar coordinate paper so dimensioned that 1.098 inches would
equal some multiple of the candlepower, then the lumens could
be determined directly by measuring the distances ab, cd, ef., etc.,
in inches and multiplying by the value of the multiple referred to
above.
Since these constants refer to the relation between the candle-
power scale of the distances from the vertical to the intersection
of the midzone radial lines and the polar curve, it follows that for
10
THE ARMOUR ENGINEER
[Vol. 4, No. 1
a certain design of polar ccjordinate paper one may construct a
scale of convenient size and shape and graduated according to
tlie above relation aiid with the common scale values indicated,
whereby the flux in lumens may be determined for any polar
curve on that design of paper. Such a construction is indicated in
Kig. 4, which shows the polar flux paper and polar flux scale as
designed by Mr. Macbeth. It will be seen that this scale (^" by
5/4"), has eight sections to correspond to eight diiTerent values
of candlepowcr per division en the polar paper for which the
46" 55'
The Armour i-n^ineeri
Fig. 5. A Simple Metliod of Determining Light Flu.x
and Splierioal Candlepower.
scale is designed. By choosing the proper section of the scale
the distances from the vertical to the points a, h, c, etc., on the
polar curve, measured by means of the scale, will equal approxi-
mately the number of lumens embraced by the corresponding
zones. By continuing this process for the entire polar curve and
adding the results the total flux will be obtained.
Still another application of the constant 1.098, as conceived
by the writer, is indicated by Fig. 5 (37), where the radial line OK
is drawn making an angle XOK (24° 3') the secant of which is
1.098. The spherical surface is divided into ten degree zones and
Jan., 1912] BARROWS: ILLUMINATION CALCULATIONS 11
the average zonal values of the candlepower represented by the
radial lines, Oa, Ob, etc. If these midzone values be projected
vertically onto the line OK and the extremity of each projection
be continued around O along the arc of a circle into the horizontal
the value of the flux in lumens in any zone will be represented by
the corresponding distance, to the same scale as the polar curve is
plotted, along the horizontal OA' measured from 0. As an ex-
ample of the manipulation of this method consider the zone be-
tween the angles of 40° and 50° from the vertical. Assume Oe,
1^
^-
-^
~ ^^
~
---
---
■-
' \
\
s
^**>v
~~
.^
1
^ \
\
\
\
\
\
\
\
\
^
\
^
^^
">*>^
^^
-
1
\
■^
^
*s^
\
\
\
\
\
\
\
^
^
^
!v
\
1
\
\
\
\
\
\
\
\
\
N
\
s.
1
1
1
1
I
\
\
\
\
\
\
\
\
\
\
\
\
s
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
11
\
\
\
85'
^-^rs"
65"
55"
45'
5" iS"
Fig. 6. Auxiliary Diagram Applicable to Any Polar
£5" SS"
The Armour Engineer.
equal to 78, to be the average intensity in this zone. Its projec-
tion on OK is equal to Oe and this value transposed to OX will
be Oe" or 60.5 lumens. In the same way the flux in the 50° — 60°
zone will be 77 lumens. By adding the lumens in the different
zones together the total lumens in the lower hemisphere may be
obtained. The flux in lumens in the upper hemisphere may be
found by projecting the midzone values upon the line OK or upon
another line 24° 3' above the horizontal, and transposing to the
horizontal as above.
The simplicity of this method is manifest. The only appa-
12 THE ARMOUR ENGINEER LVoI. 4, No. 1
ratus necessary is a pencil and piece of paper and tlie only calcu-
lation is simple addition of the nine or eighteeii values obtained.
A right-angled triangle will be found useful in securing the pro-
jections upon the line OK, but a corner of the sheet of paper on
which the curve is plotted may be conveniently substituted.
It is obvious that this method may be applied to any distribu-
tion curve plotted upon polar coordinate paper. In this case a
series of vertical lines together with the radial line. Fig. 6, laid on
a quadrant of transparent celluloid, tracing cloth, or similar trans-
parent material will greatly facilitate operations.
Since the secant of 24" 3' is approximately 1.1 (1.098) the
midzone values may be jirojectcd directly onto the horizontal line
and the sum of the projections multiplied by 1.1, and the same
results obtained as above.
The mean hemispherical candlepower may be found by di-
viding the lumens in the hemisphere by 27r (6.28) and the mean
spherical candlepower bv dividing the total flux of the lamp
by 47r (12.56).
The mean spherical candlepower of a source may also be
found by multiplying the horizontal candlepower by the spherical
reduction factor provided the mean spherical reduction factor for
that type of lamp is knowii. and the lumens fnmifl by nmltiplying
this product by Air.
Thus
F = 47r//,.
where F = number of lumens,
/ = spherical reduction factor,
/h ^ horizontal candlepower.
Having now become familiar with methods of obtaining the
amount of light available for illuminating purposes we may now
proceed with the flux of light method of performing illumination
calculations for determining approx'mately the number of lamps
necessary for installation. It consists in calculating the flux
of light in lumens available for illuminating purposes and the
amount of flux on the working plane necessary to give the desired
illumination and equating the two results. We have already
learned that the total lumers derived from a lamp is A-wIm^, where
/„,s is the mean spherical intensitv. Assume, that, due to the type
of reflector, absorption, and redirection of the rays, etc.. only a
part k of the light reaches the working plane. Then we will have
from one lamp A-n-kF^n^ lumens available for illuminating purposes.
To illuminate an area of S square feet with an average intensity
of £o foot-candles will require SB,, lumens.
Jan., 1912] BARROWS: ILLUMINATION CALCULATIONS 13
Thus the nnmher of lamps required will be
N =
4^klras K'I„
By similar reasoning the area illuminated to an average inten-
sity of £„ foot-candles by one lamp will be
In order to employ the flux of light method for practical calcula-
tions it becomes necessary to know the values of Bo for different
classes of service.
«
/.e
S/.2
'r
iae
1
lo
-O.Z
-0.4
-0.6
-o.a
1
1 1
.
-—
="
^
''
/^
^
i
/
/
^
?5
1
/
/
/
3Z 3* 3ti 3a
Relation of Sensations to Stimuli.
In determining this intensity of illumination E^ for a certain
interior it is extremely important that one possesses a clear con-
ception of Fechner's law of vision. Briefly this law states that
the sensations produced by the optical nerves vary approximately
as the logarithm of the values of the stimuli producing those sen-
sations.
This law is presented graphically by Fig. 7 (12), which is
plotted with intensities in foot-candles as abscissas, and the values
of the logarithms of these intensities as ordinates. Referring to
this curve it will be seen that the same percentage change in inten-
sity will produce the same change in sensation. Thus by increas-
ing the intensity from two to four foot-candles the same change
THE ARMOUR ENGINEER [Vol. 4, No. 1
in sensation will be effected as if the intensity were increased from
four to eight, or from twenty to forty foot-candles, the percentage
increase being the same in all cases.
A study of this law will reveal the reason for the statement
often seen in the technical press, that the effects produced by the
use of additional lamps did not warrant the additional expenditure
of energy. Thus it will be seen that every man who pretends to
handle illumination problems with engineering intelligence should
possess a practical knowledge of this fundamental law and install
a number of lamps sufficient only to enable the details of the sub-
jects illuminated to be clearly and easily perceived.
If we have given the values of the illumination intensity for
a particular class of service and the effective lumens per watt of
the equipment chosen, the determination of the number of lamps
for a particular class of lighting becomes a simple matter.
Thus in the preceding expression
AirkU, = IV K,
where W is equal to watts or cubic feet of gas per hour per lamp
and K is equal to the effective lumens per watt or per cubic foot
of gas per hour. Thus
N = ~,
WK'
S being the area in square feet, E^, the average foot-candle inten-
sity or lumens per square foot, and N the number of lamps. As
an example of such a calculation, find the number of lamps neces-
sary to illuminate a store 30 by 60 feet, or an area of 1800 square
feet, with an intensity of 3.75 foot-candles.
If the walls and ceilings are assumed to be light in color, then
with 100-watt tungsten lamps and clear prismatic reflectors there
should be obtained 4.5 lumens per watt. Thus by substitution
we have
-- SEo 1800X3.75
iV=7777P = — = 15 loo-watt lamps.
WK 100X4.5
Another method of arriving at the same result is to calculate
the watts per square foot or cubic feet of gas per hour per square
Jan., 19121 BARROWS: ILLUMINATION CALCULATIONS 15
foot, and determine the number of lamps by dividing the total
watts or cubic feet of gas required for the installation by the
amounts taken by one lamp. Thus the watts per square foot, or
cubic feet per hour per square foot are equal to the intensity in
foot-candles or effective lumens per square foot, divided by the
effective lumens per watt or per cubic foot per hour from the
luminous source, or
In this way the solution of the preceding proljlem is
w= j" = — — = 0.833 watts per square foot.
^ 4-5
The number of lamps will obviously be equal to the product
of the area and the watt^ ]ier square foot divided by the watts
per lamp, or
_ area X watts per sq. ft.
W per lamp
By using 100-watt lamps and substituting we will get as
before
^Y_i^°oXo- 833
15 loo-watt lamps.
By similar reasoning it will be found that 25 60-watt or 10
150-watt tungsten lamps will meet the requirements.
In order to be able to use these equations it becomes neces-
sary to know the working values of the constants involved. To
give any fixed value of the illumination for any class of service
or to determine the value of the effective lumens for any class of
illuminants which will hold for all conditions and classes of arti-
ficial illumination is obviously impossible. The intensity of light
required involves the physiological and psychological characteris-
tics of the individual: it depends upon the distribution and dif-
fusion of the light and the location of the source as well as upon
the amount required for the particular class of service. The
amount of effective light varies with the illuminant, with the
equipment, and with the surroundings. In the successful appli-
cation of the flux of light method much depends upon a judicial
16
THE ARMOUR ENGINEER
[Vol. 4, No. 1
study of the conditions and the surroundings. In regard to illu-
mination in general a careful study should be made of the neces-
sary and most satisfactory intensity, quality, and distribution of
light concurrent with the general welfare of the people for whose
use the system is designed. Too little light is likely to cause a
strain on the retina, while, on the other hand, too much light
should be avoided since the contraction of the iris is limited and
a strain will likewise be imposed upon the optical system.
However, values of intensities for different classes of service
which may form the basis of the flux of light method in this re-
spect, are given in Table 1. These should be considered as apply-
ing to the general and average cases, for the writer is well aware
that either from choice or necessity occasion may arise where
values far different from these should be employed.
TABLE 1.
lUuminination Intensity in Foot-candles or Lumens per Square
Foot which may be used as representing the average requirements
for the various classes of service: —
Armory or drill hall 1.5-2
Art gallery walls 3-5
Auditorium 1-3
Automobile showroom ...3-5
Ballroom .2-3
Bank 3-4
Barber shop 2-4
Billboard 4-6
Billiard room 0.5-1
Billiard table 4-S
Book-keeping 3-5
Bowling alley 1-1.5
Bowling — pins 4-5
Cafe 2-4
Carpenter shop 2-4
Cars — baggage 1-1.5
day coach 2-3
dining 2-3
mail 4-6
Pullman 2-3
street 2-3
Courts — handball 5-8
tennis 5-8
Court room 2-3
Church 2-3
Depot 1.5-2
Desk 3-5
Drafting room 5-8
Factory —
general illumina-
tion 1-2
bench illumination.4-6
Complete illumina-
tion 4-5
Fire station, at time of
alarm 2-3
at other times 1-1.5
Foundry 2-3
Garage 1.5-2
Gymnasium 2-4
Hospital corridor .0.5-1.5
wards — general .1.5-2
wards with local
illumination . .0.5-1
operating table. .8-10
dining room 2-3
bed room 1-2
lobby 1.5-2
parlor 2-3
writing room 2-3
Hotel— Corridor 0.5-1
Laboratory 3-5
Laundry 1-2
Library — stack room 1-2
Jan., 1912] BARROWS: ILLUMINATION CALCULATIONS 17
Library — reading' room, no
local illumina-
tion 3-4
reading room,
with local illum-
ination 0.5-1
Lodge room 2-3
Lunch room 2-3
Machine shop — general ...1-1.5
Market 2-3
Moving picture theatre .... 1-2
Museum 3-4
Oi^ce ...3-5
Office — general illumina-
tion 1.5-2
Pattern shop 3-4
Pool (see Billiard).
Power house 2-3
Postal service 4-6
Press room 3-5
Public square 0.5-1
Reading — good print 2-3
tine print 3-5
Residence — porch 0.2-0.5
hall 0.5-1
reception room. 1-2
sitting room. ..2-3
parlor 2-3
library 2-3
dining room . . 1-2
music room . . .2-3
kitchen 2-3
pantry 2-3
laundry 1-2
bed room 1-2
bath room . . .2-3
furnace room.0.5-1
store room.. 0.5-1
Restaurant 2-4
Rink — skating 2-3
Rug rack 10-15
School — class room 2-3
study room 2-3
assembly room. . 1.5-2.5
office 2-3
cloak room 0.5-1
[ corridor 0.5-1
I manual training ..3-5
laboratory 3-5
drawing 4-6
drafting 5-7
Sewing — light goods 4-5
dark goods 8-10
Shipping room 1-2
Show window —
light goods •. .5-20
medium goods 10-30
dark goods 20-50
Sign 4-6
Stable 0.5-1.5
Station — railroad 1-2
Stereotyping 3-5
Stock room 1-2
Store — art 4-5
baker 2-4
book 2-4
butcher 2-4
china 2-3
cigar 2-4
clothing 4-6
cloak and suit 4-6
confectionery 2-4
decorator 2-4
drug 3-5
dry goods 4-6
florist 3-5
furniture 4-6
furrier 4-6
grocery 2-4
haberdasher 3-5
hardware 3-5
hat 4-5
jewelry 4-5
lace 3-4
leather 4-5
meat 2-4
men's furnishings . .3-5
millinery 4-6
music .,.3-4
notions 3-4
piano 4-5
post cards 3-4
shoe 3-4
stationery 3-4
tailor 4-6
tobacco 2-3
Street — business 0.4-0.6
residence 0.1-0.2
prominent r e s i -
dence 0.2-0.4
Studio 4-5
Telephone exchange 3-4
Theatre — lobby 2-3
auditorium 2-3
Train shed ..0.5-1.5
Typesetting 5-10
Warehouse 1-1.5
Wharf 1-1.5
18 THE ARMOUR ENGINEER [Vol. 4, No. 1
111 regard to the effective light available from the different
types of lamps, no very exhaustive or conclusive data appears to
be available. In Table 2, however, are given values of the effective
lumens per watt for some of the electrical equipments obtained
by averaging the results obtained from various sources, and the
effective lumens per cubic foot of gas consumed per hour as com-
piled by Mr. Macbeth.
TABLE 2.
Effective Lumens per Watt or per Cubic Foot of Gas per Hour.
Gas Lamps (Small room, light ceiling).*
Cc
msumptioi
1
Effective
Type of Lamp
Glassware
cu. ft.
per hr.
Walls
lumens
p. cu. ft.
p hr.
Inverted lamp
Prismatic or opal
(concentrating)
3.33
light
dark
125
115
Inverted lamp
Prismatic or opal
(distributing)
3.33
light
dark
110
100
Inverted lamp
French roughed
ball globe
3.33
light
dark
95
70
Inverted cluster,
four .mantles
Alabaster globe
13.0
light
dark
85
64
Inverted arc,
five mantles
Alabaster globe
16.6
light
dark
87
65
Upright arc,
four mantles
Opal reflector
20.0
light
dark
75
55
Upright arc,
four .mantles
Alabaster globe
20.0
light
dark
66
48
Gas Lamps (Large room, light ceiling).**
Consumption Effective
Type of Lamp Glassware cu. ft. Walls lumens
per hr. p. cu. ft.
p hi;
Inverted
lamp
Prismatic reflectors
(concentrating)
3.3
light
dark
140
128
Inverted
lamp
Opal reflectors
3.4
light
dark
120
109
Inverted
lamp
Roughed balls
3.3
light
dark
101
75
*I'rocfeaiii.!4-s of tlic Ainci-ic-an G.ms Institute, Voluiiio 4, pajre 305, lOOi).
**Ti'aiiSiU'ti()ns of the Hliiiiiiuatiuu Euiiineeriii.i,'' Society, Volume 4, i)af::e 8()4
in09.
Jan., 1912] BARROWS: ILLUMINATION CALCULATIONS 19
Electric Lamps (Large room).
Type of Lamp
Tungsten
Tungsten
Tungsten
Tungsten
Tungsten
Tungsten
Tungsten
Tungsten
Tungsten
Tungsten
Tungsten
West. Nernst
West. Nernst
Moore light
Moore light
Arc lampt
Gem
Gem
Carbon filament
Carbon filament
Carbon filament
Carbon filament
Glassware
Clear prismatic
Clear prismatic
Clear prismatic
Satin prismatic
Satin prismatic
Satin prismatic
Enamel
Enamel
Silvered glass
None
None
Opaline globe
Opaline globe
None
None
None
Clear prismatic
Clear prismatic
Clear prismatic
Clear prismatic
Opal dome
Opal dome
Ceiling
light
light
dark
light
light
dark
light
lig'ht
light
light
light
light
light
white
light
light
light
light
light
light
light
light
Effective
Walls lumens
pr. watt
4.5
light
dark
dark
light
dark
dark
light
dark
light
light
dark
light
dark
light
medium
medium
light
dark
light
dark
light
dark
3.8
2.8
4.0
3.2
2.6
3.4
3.0
6.1
3.3
2.1
3.2
2.6
2.7
2.1
2.0
2.2
1.8
1.8
1.5
1.7
1.4
Having found the number of lamps it next becomes necessary
to determine their location. One method of doing this is to divide
the room into a number of equal areas and place an outlet over the
center of each area. If the area is of such dimensions that it can
be divided into n number of equal squares and a lamp placed at
the center of each square, then the distance d between lamps
will be
\ n
If the area be divided into n equal rectangles and a lamp placed at
the center of each rectangle whose dimensions are b and c, then
b = S -^ cN, the distance in one direction, and c == 5" -^- bN, the
distance in the other direction. With 25 60-watt lamps we would
have 72 square feet per lamp ; with L=^ 100-watt lamps, 120 square
feet per lamp, etc. Thus with 100-v.att lamps the best arrange-
ment will be as shown in Fig. 8, where the room is divided into
sections 10 feet by 12 feet, and one lamp placed over the center of
each section.
' ive- ampere, direct cm-rent eiiclosetl; clear outer, ami oi
20
THE ARMOUR ENGINEER
[Vol. 4, No. 1
The absorption of light method (18) of performing illu-
mination calculations as suggested and developed fundamentally
by Dr. McAllister promises possibilities as a means of solving
some of the intricate problems of illumination. The theory of this
method is that the lighting units within a room must produce the
sum of the lumens absorbed by the various surfaces. From this
relation one can readily determine the total lumens which must
emanate from the luminous source to produce certain incident
illumination, since for a given surface there is a direct ratio be-
tween the number of lumens absorbed and the number of lumens
reflected. From the foregoing it will be seen that the lumens ab-
sorbed by a surface is equal to the product of the incident illu-
Thc Armour Engineer,
rig. 8. .\rrangenient of Lamps.
mination, the coefficient of absorption, and the area of the sur-
face, or
where P is the lumens absorbed, a the light absorption coefficient,
Bo the average foot-candle intensity and 5" the areas illuminated.
The application of this method may be more clearly shown by
means of an example.
Assume a room 15 feet in width. 20 feet in length, and 10
feet in height, having a white ceiling, light walls and a dark floor
to be so lighted that we have an average illumination intensity of
2 foot-candles on the ceiling, 1 foot-candle on the walls and 4 foot-
candles upon the floor. Assume also that the light absorption co-
efficient of the ceiling is 0.20, of the walls 0.40, and of the floor
0.90, and determine the candlepower necessary to produce the
desired results.
Jan, 1912] BARROWS: ILLUMINATION CALCULATIONS 21
Then by means of the above formula the lumens absorbed by
the various surfaces will be
.20X2X300= 120 for the ceiling,
.40X1X700 = 280 for the walls,
.90X4X300= 1080 for the floor.
This gives a total of 1480 lumens which would require a
luminous source of 2ZS mean spherical candlepower.
It will be interesting to note in the above example that while
only 1480 lumens are generated the total efit'ective lumens on the
various surfaces due to reflection and counter-reflection will be
2X300 = 600 for the ceiling,
1X700 = 700 for the walls,
4X300= 1200 for the floor.
Or 2500 effective lumens.
If we assume that the light received by the walls from the
ceiling and floor is equal to that reflected from the walls to the
ceiling and floor and the amount absorbed by the walls supplied
entirely by the source, then by means of the flux of light method
one may very easily calculate the amounts of light which must be
directed toward the floor and ceiling in order to give the desired
intensities on these two surfaces. In the above example the total
lumens effective at the ceiling is 600. Of these 120 lumens are
absorbed and 480 lumens are reflected to the floor. Of the total
incident illumination of 1200 lumens on the floor 1080 lumens are
absorbed and 120 lumens reflected. Of the 600 lumens at the
ceiling 120 come from the floor, leaving 480 to be supplied by the
illuminants. Similarly of the 1200 lumens at the floor, 480 come
from the ceiling, leaving 720 lumens to be supplied by the lamps.
Thus we have
Efifective or Co-efficient Lumens
Area sq. ft. Intensity, incident of absorp- absorbed,
lumens. tion.
Ceiling 300 2 600 .20 120
Walls 700 1 700 .40 280
Floor 300 4 1200 .90 1080
2500 1480
Lumens received Lumens sup-
Lumens reflected. by reflection. plied by lamps.
Ceiling 480 120 480
Walls 420 420* 280*
Floor 120 480 720
1020 1480
♦Assumed.
22 THE ARMOUR ENGINEER [Vol. 4, No. 1
The discussion, thus far, on calculations, has dwelled on the
determinations of illumination intensities due to the luminous
source.
It is often desirable to know the distribution of light from a
luminous source which will give an approximately uniform illu-
mination. In order to obtain uniform illumination from oue lamp
recourse is made to the expression
Eh = ~Xcos3a,
which is transjjosed into the form
T - ^h^'
cos^ a
where Bh is the horizontal illuminination in foot-candles, a the
angle which the luminous rays make with a vertical through the
source, h the candlepower of ithe source of light a degrees from
the vertical, and h the height of the lamp above the working plane.
For uniform illumination £ and Ji., will be constant and /a at
various values of a must vary inversely as cos^ a. A polar curve
(24) showing this relation is given in Fig. 9. which represents the
distribution of intensity of a lamp in a vertical plane which would
uniformly illuminate the area beneath it. Obviously, the area thus
illuminated by one lamp is limited. In order, therefore, to uni-
formly illuminate larger areas a number of lamps must be em-
ployed and so arranged and with polar curves of such shape, as
to produce the desired effect. In an interesting article by Mr.
Wohlauer (24), problems of this nature were discussed and polar
curves derived showing various ways of obtaining uniform illum-
ination. To simplify matters we will consider first the space be-
neath and between two lamps A and B and study the distribution
of light in a vertical plane through the centers of the lamps.
The simplest way of effecting this is indicated by Fig. 10.
There the vertical candlepower of each lamp is of sufficient value
to give the desired intensity £„ beneath the source, and of suffi-
cient value in other directions that the illumination decreases in
a straight line to zero beneath the other lamp.
If we leit d = distance between lamps,
d' = distance from the lamp to point in question,
h = height of lamp,
Bo = desired illumination,
a = angle of ray with vertical,
/a = candlepower a degrees with vertical.
Jan., 1912] BARROWS: ILLUMINATION CALCULATIONS 23
then it may be shown ithat the intensity at d' due to the source,
since
£o^ d
E~d-d
7 and d = h tan a,
£,a = -T/- [d —h tan a) = ;
o
^amyt,
/o
zo
JO 40 C/O
-f-~-
1 1"^^'
c^
~j—LJeol^
1
^
^r%^
V
s
/^\
3
-^
-'•'X »
^\^
A
o" _/^
i3
?^-'^\
\\^\
/i 1
c
r/ia/4r
Fig. 9. Polar Distribution Curve for
■Uniform Illumination (One L.amp).
therefore, the equation for this curve is
P/d'-htsina\
a\ d' I
EqH^ /d' — h tan a>
cos^
This is the simplest form of curve for uniform ilkimination with a
number of lamps. It is evident, however, that with four lamps
placed at the corners of a square the illumination along the sides
of the square will be uniform, but not so at the intersection of the
diagonals of the square. The illumination at this point is four
24
THE ARMOUR ENGINEER
[Vol. 4, No. 1
times the intensity at a distance equal to V2d'\/2 from a point be-
neath one lamp, or 1.17 £o- In this figure it may be seen that the
raitio flf^-// = 1. It often happens that it is impractical to locate
lamps according to this relation. If the ceilings happen to be
low and the illuminants of high candlepower the lamps nnist l)c
placed further apart.
Under these conditions the polar curve must be made up of a
combination of those shown in Figs. 9 and 10, i. e., each lamp
must uniformly illuminate a section of the area beneath itself from
which the illumination may then assume a constant decline reach-
A
90'
a
m
^^
hif'
3^:^^
1
1
^
^
^mi
V
^
^
^
jmi
3
-^
[i\/.
/5^
\hr~
4-
/
Y^
?^yV
^^^
Co
0
^
V
^
^
X^
c
^
^
1
\^
o
1
k
3 4-
FiR. 10.
•<>lar Curves for Uniform Illuminat
(Two Lamps).
ing zero value where the uniform illumination due to the next
lamp begins. The curves of this nature are shown in Fig. 11.
Another form of polar curve for uniform illumination is shown
in Fig. 12. The equations of these curves are too complicated
for practical purposes.
The general case of polar curves (24) yielding uniform
illumination is indicated by Fig. 13. The equation for curves of
this nature is
- Id—h tan a , . Airh tan
/ol ^ \-c sm ~ —
'«(^
-").
Jan., 1912] BARROWS: ILLUMINATION CALCULATIONS 25
where c is a constant which must be determined for each par-
ticular case.
In practice the uniformity of illumination can be obtained by
choosing lamps and reflectors which direct the light in the desired
manner. Many reflectors have been designed to distribute the light
so as to accomplish this purpose. In Fig. 14 is shown the distribu-
tion of light in a vertical plane around a 100-watt tungsten lamp
when ecjuipped with an extensive type and an intensive type of
prismatic reflector. These curves show the candlepovver in radial
directions only and give no indication of the distribution of ilium-
^^M~4^
/
u^^^^c^
\\$^X^S
2
^W
4
~m5^%^
S
'P^^^wv^
S
\.,^^ t
^^
t
^\.,^
^.^
^
^
r
^y^ 1
\
^
Ftet
Fig. 12.
Polar Curves for Uniform Illumination (Two I^anips).
ination. This distribution on a horizontal plane may be calculated
from these curves by use of the equation
„ /a cos^ a
£h = T..
The results of such calculations from the curves shown in Fig. 14
are shown in Figs. 15 and 16 (12).
Curve A, Fig. 15, shows the distribution of illumination on a
horizontal plane due to one lamp placed eight feet above the plane
and equipped with an intensive reflector.
26
THE ARMOUR ENGINEER
[Vol. 4, No. 1
The ordinates are in foot-candles and the abscissce indicate
distances from a point directly beneath the lamp. The illumina-
tion is far from uniform. However, if two lamps are placed at
the same height and ten feet apart, the illumination on the plane
between points beneath the two will be nearly uniform as indicated
by curve B. Fig. 15. It will be seen that the ratio of the distance
between lamps to their height above the reference plane is 10 to 8,
or 1.25 to 1. Thus, if this ratio is maintained when the lamps
are suspended at other heights, the relative distribution of illum-
ination will remain the same, although the intensity will be some-
what less. Similar curves for lamps with the extensive type of
lar Curves,
IS'
The Armour En^it
»nd E Types of Reflectors.
reflector are shown in Fig. 16. Here the height is six feet, and
the distance between lamps is eleven feet. In this case the value
of d-^h= 11^6=1.83.
Since the general distribution of light from lamps of other
types and of other sizes is similar to these when equipped with the
same type of reflector it will be seen that the constants 1.25 and
1.83 given above refer to the type of reflector, and may be applied
to installations employing any of the commercial types of lamps,
when equipped with these types of reflectors.
The relation of the distance between lamps to the height
above the working plane d-^h is of importance and deserves con-
Jan., 1912J BARROWS; ILLUMINATION CALCULATIONS 11
sideration. The area which can be nniforml)- ilkiminated is hm-
ited in practice.
This area, of conrse, depends upon the height of the lamps,
but this, in turn, changes the intensity of illumination in the case
of two or four lamps, 'as in the examples just cited.
In Fig. 17 are shown four theoretical curves (24) for uni-
form illumination having values of K = d~h =0.5, 1.0, 1.5, and
_^
___
1"
—
B
N
\
\
% = %=I.Z5
•E
\
\
IX
<
^
^
^-^
—
.
,
Distance in Feet
16 IS 20
The Armour En^ina€r,
Fig:. 15. Illuniination Curves I Type Reflector.
-^
"^
'
^
=q
^-~
B
^
•^
^.r
\
f
\
% =
% =/.B3
\
'
r
\
0.5
^.
-~-
-J
0
i.
A
t
-
t
?
/
j
3
/
#
/
5
/<
9
ZO
Fig. 16. lUui
Distance
ation C'ui
' Faer.
Tht
E Type Reflector.
Curve A. Illumination Due to One Lamp.
Curve B. Illumination Due to Two Lamps.
2.0. In this case the height and illumination are the same. It will
be seen that the ratio of d^x = 2.0 is as great as is practical, and
that in most cases this relation will be much less. This expression
d^i gives the minimum height h at which the lamps must be
placed when located a distance d apart. It can be shown that the
same units equally spaced can be raised above the minimum height
without impairing the uniformity of illumination, while a suspen-
28
THE ARMOUR ENGINEER
[Vol. 4, No. 1
sion lower than the minimum will result in non-uniform illum-
ination. For one lamp it is evident that the illumination will be
uniform beneath it regardless of the height, provided the distrbu-
tion of hght is favorable, and that the intensity will vary inversely
as the height, as shown by the curve RS7\ Fig. 18 (26).
If two lamps are considered the conditions existing will be as
shown on the right of Fig. 18. For uniform illumination along a
line connecting points beneath the lamps the minimum height is
four feet, as shown in the right of figure. At a height of two
feet the illumination reaches zero between the two lamps, while
for heights greater than four feet the uniformity is unimpaired,
o
1^^^=^--LI1|| S<7 ••
10
^^^^r
5^
™^^^^^^\r~'~~-^^
r
70
^^^^^S^^^^^^^^?----^
fio
^(^t^-p^ \ .^'X <; V. J\ \ \, ~^\ ^^^
i]
^^* ^'''
4
r
%^\
/
\ \ \r\
Fig. 17. Polar Curves for Uniform Illuinination for Different Ratios of Height
and Distance.
but the intensity becomes less. The change in intensity beneath
the lamps for different heights is shown by the curve RST, while
the intensity midway beneath the two is indicated by VSU .
With two or four lamps the intensity of illumination beneath
them ceases to vary inversely as the square of the height, or dis-
tance to the surface illuminated, since the light flux reaching the
surface changes only by the change in amount wdiich passes out-
side its boundary. In large interiors with a number of lamps
which throw the light in a downward direction the intensity of
illumination on the floor or working plane will vary slightly with
different heights of suspension of the lamps ; since all of the light
from the lamps centrally located will strike the floor no matter
Jan., 1912] BARROWS: ILLUMINATION CALCULATIONS 29
what the height and the difiference in the total flux is due only to
that amount which strikes the walls from those lamps around
the outer edges, part of which will be reflected back.
The preceding discussion by no means embodies the entire
calculations pertaining to illumination. There are others less
practical and still others involving long discussions and tedious
mathematical deduction quite out of the scope of this article, al-
though of great value to the student of illuminating engineering.
That these dififerent branches of the subject may be investigated
further by those interested, the titles and references of some of
the leading articles are incorporated in the following bibliography.
For instance, the distribution of illumination in the neighlx)rhood
/f
I
,
f
1
1 *'
2
1
1
1
i
1
1
•s
1
1
1
1
s .
f
^
1 /
^ 1
4"
•s
^
Yy
N
t
^
rr
"r
/
t-
s
k
Height of Suspxnalo,
Figr. 18. Kflfeot of Height of Lamps
ination.
The Arm,ur En^mttr
Distribution and Intensity of Illi
of a row of lamps (30) illumination from linear sources (31) and
the calculation of radiation from surface sources {^i^) and (36),
refer to special branches of the subject not entered upon by the
writer.
There are also references to articles dealing with the subjects
discussed in this paper and explaining the methods more in detail
and to greater length.
30 THE ARMOUR ENGINEER [Vol. 4, No. 1
BIBLIOGRAPHY OF ILLUMINATION CALCULATIONS.
(1) Rousseau diagram.
Palayz Industrial Photometry, p. 20.
(2) Data on indoor illumination and the Rousseau diagram.
Trans. Ilium. Eng. Soc. 1, 245, March, 1906.
(3) New graphic method for determining the mean spherical inten-
sity of a lamp by length of a straight line when mean meridi-
onal intensity is given.
Elect. World, 51, 645, March 18, 1908.
(4) A rectilinear graphical construction of the spherical reduction
factor of a lamp.
Trans. Ilium. Eng. Soc. 3. 243, April, 1908.
(5) Fluxolite paper and principles involved.
Ilium. Eng., N. Y., 3, 655. February, 1909; 4, 491, November,
1909; 4. 148, April, 1909; 5, 132, May, 1910.
(6) Calculation of mean spherical candlepower.
Trans. Ilium. Eng. Soc, 3, 27, March, 1908; 4, 436, May, 1909.
(7) Graphic illumination chart.
Trans. 111. Eng. Soc. 2, 579. September, 1907.
(8) Calculation of illumination with the ordinary slide rule.
Ilium. Eng., N. Y. 3, 152, May, 1908.
(9) A "Calculator" for the use of illuminating engineers.
111. Eng., N. Y. 3, 21, March, 1908.
(10) Calculation of illumination by the flux of light method.
Trans. 111. Eng. Soc. 3, 518, October, 1908.
(11) Simplification of illumination problems through the conception
of light flux.
Trans. Ilium. Eng. Soc. 4, 310, April, 1909.
(12) Theoretical notes on interior illumination.
Elect. Rev., N. Y., 56, 793, April 16, 1910; 56, 847, April 23, 1910.
(13) Operating efSciencies of some commercial installations of light-
ing systems.
Trans. Ilium. Eng. Soc. 4, 849, December, 1909.
(14) Comparative practical efificiency of various types of gas lamps.
Trans. Ilium. En/g. Soc. 4, 96, February, 1909.
(15) Some results obtained through illuminometry.
Trans. 111. Eng. Soc. 4, 789, November, 1909.
(16) Experimental data on illuminating values.
Elect. Rev., N. Y. 5L 986, December 21, 1909.
(17) Illumination tests.
Trans. Ilium. Eng. Soc. 5, 391, May, 1910; 5, 553, October, 1910.
(18) Absorption of light method of calculating illumination.
Elect. World, 52, 1128, November 21, 1908; 55, 1388, May 16,
1910.
Jan., 1912] BARROWS: ILLUMINATION CALCULATIONS 31
(19) Coefficients of diffuse reflection.
Trans. 111. Eng. Soc. 2, 653. October, 1907.
(20) Reilection from ceilings, walls and floors.
Trans. 111. Eng. Soc. 3. 584, October, 1908.
(21) Location of lamps and illuminating efficiency.
Trans. 111. Eng. Soc. 1. 321. 1906.
(22) Check on the reliability of photometric curves.
Trans. Ilium. Eng. Soc. 2. 645, 1907.
(23) Spacing of light units.
Elect. World, 54, 663, September 16, 1909.
(24) Uniform illumination of horizontal planes.
Elect. W^orld, 50. 1207, December 21, 1907.
(25) The number of lamps for uniform illumination.
Elect. W orld, 51. 1376. June 27. 1908.
(26) Effect of height of .suspension upon uniform illuiiiiiiation.
Elect. World. 51. 601. .March 21, 1908.
(27) Engineering problems in illumination.
Trans. 111. Eng. Soc. 3, 693. November, 1908.
(28) Standard relation of light distribution.
Trans. Ilium. Eng. Soc. 4, 745, November, 1909.
(29) Distribution of illumination in the neighborhood of two lamps,
Elect. World, 47, 917, May 5, 1906.
(30) Distribution of illumination in the neighborhood of a row of
lamps.
Elect. World, 48. 805, October 27, 1906.
(31) Illumination calculations for linear sources.
Elek. Zeit. 28, 757, August 1. 1907; 29, 883, September 10, 1908.
(32) Geometrical theory of radiating surfaces with discussion of light
tubes.
Bui. Bureau of Standards, 3, 81. April, 1907.
{33) The mathematical theorv of tinite surface light source.
Trans. Ilium. Eng. Soc. 4, 216, April, 1909; 5, 281, May, 1910.
(34) Distribution of curves of radiation from plane, circular and
cylindrical radiators.
"Radiation, Light and Illumination."' pp. 190, 193, 195, 197.
(35) Suggestions for records and a system for compiling illumination
data.
Ilium. Eng. (N. Y.) 5. 126, May, 1910.
(36) Graphical solution of problems involving plane surface lighting
sources.
Elect. World, 56. 1356, Dec. 8, 1910.
(37) A simple method of determining the mean spherical candlepower
and light flux.
Elect. Rev. & West. Elect. 58, 440, Mar. 4. 1910.
THE UTILIZATION OF BLAST FURNACE GASES.
BY EDWARD K. HAMMOND.*
In smelting iron ore. according to ai.>proved practice, approx-
imately one hundred and fifty thousand cubic feet of gas is evolved
from the blast furnace for every ton of pig iron that is produced.
This gas was formerly discharged from stacks at the tops of the
furnaces, but the presence of twenty-five per cent of carbon mon-
oxide, a gas of high calorific value, attracted the attention of metal-
lurgists to the possibility of utilizing it for fuel purposes. The
first important step in this direction was made in burning the
blast furnice gases in the preheating stoves, which will be subse-
quently described, where the temperature of the air blast supplied
to the furnaces is raised to a temperature of about 930 degrees
Fahrenheit before coming into contact with the fuel. This greatly
facilitates the efiiciency of operating conditions, but the supply of
gas produced by a given number of furnaces was found to be far
in excess of that required for preheating the blast which they use.
A further possibility of progress was thus presented, and with the
application of internal combustion engines as prime movers for
power generation it became evident that the successful application
of blast furnace gases in this direction would mean an enormous
saving of fuel.
As was to be expected, in a countr}- where the natural re-
sources have been greatly depleted, the Germans were pioneers
in the application of blast furnace gases for power generation and
it is to the researches of this race of persistent investigators that
the credit for this advance in the metallurgy of iron and steel is
due. But when the practicability of this method had once been
demonstrated, the industrial companies of our own country acted
with their characteristic initiative and we now have several power
plants operating upon blast furnace gases.
By far the most important of these is the installation at the
Gary Works of the Indiana Steel Company, where the scale upon
which this principle has been applied far exceeds that of any
other plant in the world. It is worthy of notice in this connection
that, at the time when the use of blast furnace gases was decided
upon as a source of power, practically no operating experience had
been obtained in this country, where differences in ore, fuel, and
metallurgical practice combined to render the results obtained in
Europe of little value. Work had been conducted in this country
*Class of 1909. With the City of Chicago, Gas Testing Laboratories.
Jan., 19121 HAMMOXD: BLAST FURNACE GAS 33
upon an experimental scale and plants had been ordered for in-
stallation at the South Chicago Works of the Illinois Steel Com-
pany, for the Edgar Thomson Mills at Pittshurg, and for the
Lackawanna Steel Company at Buffalo. The latter installation
was completed and placed in operation in 1902, which gives it the
credit of being the first plant of its kind in the United States.
But neither of these plants were in operation at the time that the
equipment was ordered for the Gary Works, so that the courage
in their convictions exhibited by the designers is particularly no-
ticeable. Nor was this confidence misplaced, for, after two years
of successful operation, it has been found that the load can be
carried with the greatest regularity.
When completed, the Gary Works will have sixteen blast
furnaces at their command. At the present time, however, eight
furnaces comprised in the groups of 5 to 12 are in operation and
the system for the utilization of blast furnace gases refers to their
production. Furnaces 1 to 4 are now in course of construction and
a complete installation for using the gases which they produce is
being built.
Before entering upon a discussion of the methods that are
used to fit the furnace gases for use in the gas engines, a brief
consideration of the reactions that are responsible for their forma-
tion will not be out of place. For the benefit of non-chemical
readers, it may be mentioned that the blast furnace consists of an
approximately cylindrical structure, those at Gary being 88 feet
in height by 32 feet in diameter. These furnaces are lined with
fire brick. Neglecting the fluxing material that is added to remove
impurities, the charge consists of iron ore and coke which is fed
into the furnace by a car that runs up the inclined track shown in
Fig. 2. A bell-shaped valve at the top of the furnace admits the
charge as required but prevents the gases from escaping. The
pipe that surrounds the furnace at the base delivers the blast of
preheated air through openings called tuyeres. The blast then
rises through the bed of ore and coke in the furnace and leaves
through four openings at the top, that communicate with two
ducts which carry away the eas. The process by which carbon
monoxide is formed is somewhat complex but may be fairly well
expressed by the following equations. Upon entering the furnace,
the oxygen in the blast reacts with the coke, which is at a red
heat, pccorcHng to the following two equations:
C+0., = CO.. (1)
C0,+ C = 2 CO. (2)
For the sake of simplicity, we shall assume that the ore in
the furnace is a pure red hematite, which has the composition
34 THE ARMOUR ENGINEER [Vol. 4, No. 1
FcnO:,. The following reactions will then commence as soon as
the ore enters the furnace :
2 Fe,0, + 8 rO = 7 CO., + 4 F^ + C. (3 )
2Fe,0, + CO = 2FeO-\-Fe,0,-{-CO,. (4)
If these were the only reactions taking place between the ore and
the fuel at the top of the furnace, the gas given off would be car-
bon dioxide, which has no fuel value. But reactions (3) and (4)
are limited by the fact that both metallic iron and carbon have
the power of reducing carbon dioxide to form combustible carbon
monoxide according to the reactions shown in equations (5) and
(6),
Fe+CO. = FeO-^CO. (5)
C-^ CO, = 2 CO. (6)
The result is that the gases which are evolved from the furnace
have an average composition which is expressed by the analysis
;iven t
lelow :
CO
25%
Carbon monoxide.
H
3%
Hydrogen.
CO.,
129?
Carbon dioxide.
N
609?
Nitrogen.
and this gas has an average calorific value of 95 B.t.u. per cubic
foot.
The successful use of blast furnace gases in internal com-
bustion engines depends very largely upon the degree to which
their purification has been carried. This fact has led to the instal-
lation of a most complete equipment at the Gary Works for carry-
ing out the various purification processes through which the gas
passes before it reaches the dififerent points of consumption. A
good idea of the arrangement of this equipment will be obtained
from the diagram shown in Fig. 1. which presents a panoramic
view of the equipment together with data concerning the quantities
of blast furnace gas used in dififerent parts of the plant.
The gas produced in each furnace is discharged through four
openings at the top. as shown in Fig. 2, and enters two ducts
called "downcomers" which carry it to the first dry cleaner or pri-
mary dust catcher, where the bulk of the dirt and ore dust is
removed. These dust catchers are cylindrical tanks 45 feet in
height by 25 feet in diameter, with conical tops and bottoms, and.
like the downcomers leading to them, they are made of steel plate
without anv fire brick lining: the purpose is to cool the gases as
far as possible by radiation from the plates. The dust-laden gases
entering these catchers have their direction of travel abruptly
|1 .5 I !3-»R
36
THE ARMOUR ENGINEER
[Vol. 4, No. 1
changed, and at the same time, their speed is greatly reduced
owing to the increased cross-sectional area. This results in set-
tling out a large part of the dust. The gases leave these prmiary
dust catchers at the top and are carried to the secondary dust
catchers, which are of exactly the same type but smaller, their size
being 30 feet high by 18^ feet in diameter. These secondary
Fig. 2. The Blast Furnace, Dust Catchers, and Preheating Stoves.
catchers remove a further part of the dust carried by the gas and
also serve the purpose of water seals when it is required to shut
down the furnace for repairs. This is made necessary through the
arrangement of the furnaces in pairs, each furnace of the pair dis-
charging its gases through primary and secondary dust catchers
Jan., 1912] HAMMOND: BLAST FURNACE GAS 37
into a common main. When one of the pair of furnaces is shut
down, its secondary dust catcher is filled with water to a level
above the inlet pipe, thus forming a water seal that prevents the
gases from the opposite furnace from backing through it.
The preceding description of gas purification applies to the
two pairs of furnaces comprised in the group 9 to 12, which were
the first ones to be built. The same method of preliminary puri-
fication is followed with the gases from furnaces 5 to 8, but the
process is carried a step farther. In this case the gases issuing
from the secondary dust catchers pass into large settling t«nks
which are 40 feet in diameter by 25 feet high, one of these tanks
being provided for each pair of furnaces. In passing through
these tanks, the speed of the gas is so far reduced that a further
portion of the dust with which it is contaminated is allowed to set-
tle out. These settling tanks discharge the gas passing through
them into a common main, which corresponds to the one uniting
the secondary dust catchers of furnaces 9 to 12. All of the dust
catchers are equipped with bell bottoms through which the dirt
which they collect is discharged into cars that run on tracks pass-
ing under them.
In all cases, the gases now pass to the groups of primary
washers, three of which are provided for each pair of furnaces.
Each of these washers has a capacity sufficient to take care of the
gases produced by a single furnace, so that one of the group is
always available in case it is necessary to shut down a washer for
repairs or cleaning. Fig. 3 illustrates one of these washers in
course-of construction, showing that they are cylindrical in shape
with conical tops and bottoms. These washers are 40 feet in
height by 22 feet in diameter, and they are filled about one-third
full of water, this level being maintained constant by overflows.
Fresh water is fed in from pipes at the top. The gas is led into
these washers at the top and carried down almost to the surface
of the water by means of a central pipe, which is provided with a
fluted mouth to spread the gas out over the greatest possible sur-
face. In this way, a considerable amount of the dust still carried
by the gas is thrown down and absorbed by the water. After
impinging upon the water, the gas rises to the top of the washer
and leaves through two openings, one at either side, where it
passes through dust legs into the main. At this point, a division
of the gas is made, according to the data presented in Fig. 1. The
gas that is to be used for preheating the blast in the stoves and for
use under the boilers has been sufficiently purified, and the re-
quired amount is diverted for these purposes.
Eight preheating stoves are used in connection with each
pair of bla.st furnaces, as shown in Figs. 2 and 3 — four stoves for
38
THE ARMOUR ENGINEER
[Vol. 4, No. 1
each furnace. These stoves have a central flue surrounded by a
checker work of fire l)rick, and in operation the gas. mixed with
the necessary amount of air for its combustion, is burnt in the
central flue. The hot products of combustion then pass down
from the top of the flue through the checker work and then out
into the stack. This results in heating up the checker work to a
bright red heat. In using the stove for preheating the air blast,
the process is reversed. Tn this case, the air passes up through
the hot checker work, where its temperature is raised to about
930 degrees Fahrenheit, and then down through the central flue
and out into the duct leading to the furnace. The stoves are used
Fig.
The Frimar.v Wasliers, Blast lurnace, and Preheating Stoves in Course
of Construction.
in rotation, each one being heated up for a period of one hour,
after which the valves are reversed to shut oflf the gas and admit
the air blast. As shown in the diagram. 30 per cent of the gas
from the furnaces is used in the preheating stoves, this amount-
ing to 6,750,000 cubic feet of gas per hour.
The boiler settings, where blast furnace gases are used for
fuel, have been modified so that they may be used for either gas
or coal. When the plant is operating under normal conditions,
the boilers are of relatively small importance, only enough steam
being maintained to keep the four steam blowing units going, the
Jan., 1912]
HAMMOND: BLAST FURNACE GAS
39
idea being to have them available for taking up the load in case
of the failure of the gas engines. These steam blowers are also
necessary in starting up the plant to obtain a gas supply for the
blowing engines which depend upon this fuel. The boilers also
supply the two General Electric Curtis turbines in the electric
power station, the hydraulic presses used in the plant, and small
amounts of steam for other purposes. The boiler equipment con-
sists of sixteen units of 400 horsepower, which thus provides a re-
serve of 6,400 horsepower. The boilers are of the Sterling and
the Rust water-tube types, eight boilers of each kind being housed
in separate buildings. A firing bed of coal is maintained under the
■1' ^^^rn.
1 ' ^
Figr. 4. General View of Secondary Washers, Water Separators, and Gas Holders.
boilers, which is hand fired on account of the small amount of
this fuel that is used. It will be seen in the diagram that IVz per
cent of the total gas supply, or 1,700,000 cubic feet of gas per
hour, is burned under the boilers.
Returning to the portion of the gas requiring further puri-
fication to fit it for use in the gas engines, it will be seen in the
diagram that a duct conveys it from the primary to the secondary
washers. This duct is provided with a series of dust pocker-
which assist in removing some solid impurities from tlie ga-
passing through them. Four Zschocke washers are u>e.l tn trea:
40 THE ARMOUR ENGINEER [Vol. 4, No. 1
the gases from each pair of furnaces and each of them has a
Theisen washer connected in series with it. The design of the
Zschocke washers used at Gary has lieen mochfied in that they
have the checker work placed above tlie baffle ])lates in a single
vertical drum 60 feet in height. These washers will be seen at the
extreme left in Fig. 4. The building which houses the Theisen
washers is shown in course of construction in this picture, where
one of the washers is plainly visible. The water separators, which
will be subsequently mentioned, are the vertical cylinders to the
right of the Theisen washers from which the gas passes into the
main leading to the holders shown at the extreme right.
The gas enters at the l)ase of the Zscl'ocke washers and
traverses a winding course between the baffle ])lates in the lower
half of the drum. On leaving this section, the gas rises through
the checker work, where it meets a spray of water from the top.
This water is fed into the washer through Schutte & Koerting
spray nozzles which are mounted on radial pipes extending out
from a central water main, the purpose being to obtain the great-
est possible distribution. This treatment removes the greater part
of the dust which is still carried by the gas. On leaving the
Zschocke washers, it enters a building which houses the Theisen
washers, where the final step in the purification process is car-
ried out.
The Theisen washer consists of a steel cylinder which sur-
rounds a revolving drum that carries a number of vanes on its
periphery. This drum is driven at 350 r.p.m., and the vanes pick
up water at the bottom of the cylinder and spread it over the
entire inside surface of the machine. As the gas passes through
these washers it impinges upon this film of water, which takes
up any impurities which are still held in suspension.
The gas is now free from all impurities except a considerable
amount of moisture that has been taken up in the washing pro-
cesses and is carried in the form of mist. This moisture is re-
Tnoved in the water separators, shown at the right of the Theisen
washers in Fig. 4, which consist essentially of boxes containing
iron shavings. The gas strikes these shavings at high speed and
deposits most of the water which it carries upon them. It then
reverses its direction of travel, so that there is no opportunity
for the water to be picked up again. Passing on from the water
separators, in a comparatively dry condition, the gas goes into
the holders under a slight pressure. These gas holders are 90 feet
in diameter by 30 feet in height and are of the standard type.
Reference to the data presented in Fig. 1 will show that 15
per cent of the total gas supply, amounting to 3,400,000 cubic feet
per hour, is used in the gas engines that drive the blowers. These
Jan., 19121 HAMMOND: BLAST FURNACE GAS
engines are housed in two buildings, one of which supplies each
group of four blast furnaces through independent blast mains, but
a common gas main connects the two stations to provide for a
possible failure in the gas supply from one group of furnaces.
Fig. 5 shows an interior view in one of these stations, the equip-
ment of which consists of eight gas blowing units, having a total
capacity of 265,000 cubic feet of free air per minute, and two
steam blowing units, having a capacity of 45,000 cubic feet of
free air per minute. The steam units are held entirely in reserve,
and three gas units and one reserve are operated to supply each
pair of furnaces. Each gas driven blower has a capacitv of 33,000
cubic feet of free air per minute. The division of output will
thus be evident.
The gas engines have a rated capacity of 3,000 indicated
horsepower, and the steam engines 3,500 indicated horsepower.
TABLE 1.
Data on Operation of Gas Blowing Engines for Month of October,
1911. Furnaces 5, 6, 7, 8 and 11.
Av. I.H.P. Av. total c. ft. Av. c. ft. gas Coal eq. gas
Developed gas per day pr. I.H.P. hr. pr. I.H.P hr
5 •• 2,663 8.924,000 139 132
6 2,885 9,669,000 140 133
7 3,007 10.066,000 139 l'32
8 2,819 9,420.000 139 l'32
11 2,704 9,300,000 143 1.36
14,078
14,078:8=1,760 average horsepower per engine.
Cubic feet of gas given is equivalent cubic feet at 100 B.t.u.
Average B.t.u. of gas for October, 1911, was 94.5.
The general arrangement of the equipment is essentially the same
as that of the electric power station, which will be subsequently
described. The results obtained in operating these gas engines to
supply blast to five furnaces are presented in Table 1 .
An interior view of the electric power station is shown in
Fig. 6, which gives a good idea of the general arrangement of the
equipment. There are seventeen twin-tandem, double-acting gas
engines running at 83 1-3 r. p. m. Fifteen of these are arranged
for direct coupling to 25-cycle, 3-phase, alternating current genera-
tors, which furnish power at 6,000 volts. The other two gas en-
gines are arranged for direct coupling to 250-volt direct current
generators. These twin-tandem engines have a rating of 4,000
horsepower, and the generators of 2,000 kilowatts, but in order to
assure having ample power, they are designed to carry a 30 per
cent overload. Two General Electric Curtis turbines, of 2,000
42
T?IE ARMOUR ENGINEER
IVol. 4, No. 1
kilowatts capacity, have also been installed in this station for use in
case of emergency.
The twin-tandem engines were built by the Allis-Chalmers
Company and do not differ materially from their standard engines
of that type. As shown in the illustration, they are set cross-wise
of the building, where the cylinders are conveniently situated for
connection with the fuel suj^ply. The gas mains for this purpose
pass over froiu the holders at a sufificient height to clear the tracks
and connect with a main running along the wall of the electric
power station. The leads to the engines enter the building at the
Fig. 5. Interior View of tlie Blowing Engine House.
level of this main and then drop down below the floor level, as
shown in Fig. 6, where the connections with the engine cylinders
are located. The exhaust from the engines is delivered into a tun-
nel running parallel with the power station. This tunnel commu-
nicates wdlh a steel stack, 92 feet in height, which carries away all
objectionable vapors given off and also does away with any nui-
sance from the noise of the engines. It will be seen, in the data
presented in Fig. 1, that 45 per cent of the total gas supply,
amounting to 10,000,000 cubic feet per hour, is used in the gas
engines in this power station.
The saving effected through the use of blast furnace gases
Jan., 19121
HAMMDXD: BLAST FURNACE GAS
in generating power will be evident from the fact that, for every
ton of pig iron that is produced per day, 25 brake horsepower is
available for power purposes, outside of the power required to
operate the blast furnaces, provided thi-^ power is generated in
gas engines.
An idea of the magnitude in the saving for coal, at the Gary
Works, through the use of blast furnace gases for the engines in
the electric power station and the blowing engine houses, may be
obtained through reference to Table 2. The importance of this
station from an economic standpoint will be further appreciated
Fig. 6. Interior View of tlie Electric Power Station.
when it is known that all of the electric power used in the Gary
mills is but a fraction of the output of the electric power station.
The City of Gary obtains all of its electric power from this source,
the current being sent out at 6,600 volts and transformed to the
desired voltage. In addition to this. The xAmerican Bridge Com-
pany, The American Sheet and Tin Plate Company, and The
Buffington Cement Works, obtain all of their electric power over
a 22,000-volt line from Gary. And all of this power is obtained
through the utilization of a by-product, — a source of revenue
which most of the steel mills in this country allow to go to waste.
44 THE ARMOUR ENGINEER [Vol. 4, No. 1
TABLE 2.
Coal Saved Through Using Blast Furnace Gases at the Gary Works.
Total capacity of gas engines installed to date:
17 Engines at 2.500 K. W. or 3,350 H. P 56,950 H. P.
16 Engines at 2,000 iH. P 32,000 H. P.
6 Engines at 3,000 K. W. or 4,000 H. P 24,000 H. P.
112,950 H.P.
or say 113,000 H.P.
Running 24 hours per day and 30 days a month, the total horse-
power hours in one month would equal —
113,000x24x30, or 81,360,000 H. P. hours.
Taking average cubic feet of gas per I. H. P. at 130 and coal equiv-
alent per I. H. P. at 1.3, the total coal equivalent would be —
81,360,000x1.3:2,240, or 47,000 gross tons (about).
Taking efificiency of steam engine at 10 per cent and gas engine at
25 per cent, the saving in coal by using above capacity in gas engines
over steam engines would be 70.500 gross tons per month.
SYMBOLIC REPRESENTATION OF ALTERNATING
CURRENT QUANTITIES.
BY HAROLD W. NICHOLS, M. S., E. E.*
The symbolic method for the representation of alternating
current sine waves, clue to Steinmetz and others, is possible when,
and only when, the quantities of the problem have all the same
frequency. In this case, with the understanding that the fre-
quency is not ca]ial)le of variation, the simple harmonic function
f = a sin (pt — cf>)
is known when the elements a and 4> are given ; / can therefore
be represented in magnitude and phase by a single stroke in one
plane. This line is supposed to rotate with an angular velocity p,
in common with the other sinusoidal quantities of the problem, so
that their relative positions do not change with the time. The
diagram may therefore be brought to rest and there remains the
ordinary plane "vector" diagram. Points in this diagram repre-
sent only two of the three variables necessary to completely
specify the sine wave : the diagram itself is one section of the
three dimensional figure, taken at the particular frequency chosen
The possibility of application of this method is subject to the
frequency limitations just mentioned, and any attempt to make
it reach farther will, of course, lead to absurd results.
The purpose of this article is to show how the notation must
be modified in order to represent multiple frequency quantities,
such as power. I shall first show a modification of the method
usually employed, and then develop a new one which is free from
the difficulties of the old.
I.
1. The simple plane vector in the symbolic notation is
e = ^1 H- ;>o
where, with the convention of complex variable theory, / de-
notes rotation by + 90°. This vector, if it represents a true sine
function of the time, will be called a vector of type (A). Exam-
ples are waves of e. m. f. and of current.
*Class of 1908. lustriictor in Electrical Engineering, Armour Institute of Tech-
nology.
46 THE ARMOUR ENGINEER [Vol. 4, No. 1
(The term "vector" is not strictly correct when used in this
sense; the e. m. f.'< and currents in Hnear circuits are not vectors,
and the onh" \va\- in whicli direction in space is associated with
them is in the device of representing them in a plane. The
directed lines in this plane are plane vectors, but they do not cor-
respond to e.m.f. and current in a vector sense. I shall use the
.term, however, in order to conform to ordinary usage among
engineers.)
A vector, as z = r -[- /.r. will be called a vector of type (B)
when it is not a function of the time. (Impedance, admittance.)
The confusion at ])resent existing in alternating current
theory is due to treating these two vectors in the same manner.
A few examples will show how this confusion arises.
The product of two vectors
i, + //■., (A)
r-\-jx (B)
is a vector of ty]>e ( V) : in the case considered, (usual notation),
it is the e.m.f. drop across r+/.r. We may formulate this result
(A) (B) = (A)
The product of two vectors
e,+je, (A)
h+jh (A)
i<^ not a vector in the complex plane, since it is of double fre-
quencv. It is true that formal multiplication gives a quantity
of the form P -{- jO. hut since the frequency has changed the
operator / has no longer the same meaning. This is the reason
that the product does not represent the power in the circuit sec-
tion considered, but gives an average value equal to zero. To over-
come this difficulty. Steinmetz changes the laws of operation with
/, so that in the case of a double frequency quantity operations
with / and unity are not commutative. A consideration of these
two examples shows the necessity for distinguishing between
the two types of vectors.
2. Let vectors of t}pe (A) be distinguished by the prefix a,
thus ' '^ ">
e = o-(^, +;>.) (A)
where o- is an operator satisfying the further conditions
(7J = —ja,
a-^j = X.
Jan., 1912] NICHOLS: SYMBOLIC REPRESENTATION 47
This is in effect the device hitherto employed to overcome
a fundamental difficult)- inherent in the complex method, al-
though the introduction of the additional operator is an innova-
tion.^ This system of definitions is not elegant, considered as the
starting point for a calculus of operations : the most that can be
said for it is that it "works."
The operator a, applied to a vector, as /, + ji.-,, imparts to
it a constant angular velocity p. Xow suppose that this A-vector
is to be multiplied by a B-vector, say z = r -\- jx, a time constant.
By the ordinary rules of complex algebra this means to multiply
the scalar values of (A) and fB) and add their phases, and the
question now arises, — at what time are the phases to be added?
The answer is as follows : The A-vector has for its complete
representation
i = /exp (;>f+/>)
where 4> is the phase difference between the rotating vectors
exp (//>/) and i. Similarly,
z = Z exp (jo + /a)
and a comparison of these shows that the phases are to be added
at the time ^ = 0 (or 2^11 :p). if the result is to be represented in
a stationary diagram, and without changing the position of z.
Hence we get the peculiar result that the / in exp (jpt) is not the
same as the / in exp ( j<f)) except in the trivial case, t = 2iTn:p.
To indicate this limited aspect of exp (jpt), I have denoted
it by a, so that our operator shows the functional dependence of
the A-vector upon the time. The only time that combination of
this operator with / is possible is at the time t = lirfi-.p.
The operation aa means sim])ly that the product A A is no
longer of fundamental frequency, and hence cannot be repre-
sented in the complex plane.
Evidently |o-^| = 1, and if we take a-j = X, then A may be
defined as the analogue of / in another complex plane of order
two. typified by the double frequency operator a". The meaning
of this operator will develop later.
3. Now consider the product of two vectors
a (i, -\- ji.-.) (A)
r^j.v . (B)
it is
e =o- [ f r/\ — xl, ) -\- j( ri. — xi, ) ]
which is a vector of type (A), as it should be, and represents
the e.m.f. over the circuit section r + /-i'. This is shown by the
48 THE ARMOUR ENGINEER [Vol. 4, No. 1
operator a. Its appearance as a first ])o\\er indicates that the
vector is of fundamental frequency.
Take now the product
e = a{c, + jc,) (A)
i = ,r(j,+jL) (A)
it is
P = a- ( r,/, + .rt','V) + <T-j (>,/, — ^ij
which is the sum of a constant and a douljle frequency scalar and
represents the power in the circuit. The tirst term is the real
power, and the second the "wattless" part, as indicated by the
prefix A. This may be taken as the definition of A.
Power expressed in any other way, as i-z, evidently gives
the same result, since it is the product of two A-vectors and a
B-vector which does not ct)ntain the operator a.
The quotient of an A-vector by a B-vector is an A-vector,
since the B-vector does not contain a.
The quotient of two A-vectors is a B-vector, ]j>- formal
division.
Difit'erentiation of vectors of either type by scalar variables
leaves the type of the vector unchanged.
4. It has been shown that the introduction of another
operator, as a, characterizing vectors of type (A), i> necessary
to clear up the confusion in connection with "'vector power."
This confusion is due to grouping together vectors of the two
types without making the proper distinction between them.
The operator A is not associated with direction in the plane
of reference, since the operation an- takes the product out of this
plane, — it indicates a double frequency quantity of scalar char-
acter. The ambiguity in the sign of A, depending ujjon the order
in the product ei, depends upon the direction of circulating
power in the circuit, and is immaterial for our purpose.
II.
The calculus of operations whose basis is the set of defini-
tions already given is not satisfactory, either for operating with
alternating current vectors, or for clearly understanding them.
The difficulty lies in the fact that certain operations lead to a
quantity which cannot be interpreted in the complex plane, since
it is not of fundamental, but of double, frequency. The power in
an alternating current circuit is the sum of a constant scalar part
c,/\ -|- eJn, the average power, and a wattless, circulating part of
double frequency, and to represent the two a very artificial device
Jan., 1912J NICHOLS: SYMBCJLIC REPRHSHNTATiON 49
is employed, which I have given analytically in the first part of
this paper. The alternative method which I propose follows :
Three distinct kinds of quantities are considered in the alter-
nating current circuit (sine waves only, i. e., not distorted waves,
assumed), characterized by frequency, — those of zero, first, and
second order. Xovv consider three mutually perpendicular axes
in space ; let one be the axis of reals, 1, the second the axis of
pure imaginaries, j. and the third that of another double fre-
quency unit which I will call k. Then fundamental frequency
vectors will lie in the ( l.j ) plane, and double frequency in the (l,k)
plane, that is, parallel to k. Any two vectors, e or i, of the fun-
damental frequency, will be represented by their components,
thus :
e = 1 £-, + j r, + ko
i = 1 ■/, + j L -f ko
since they lie in the (l.j) plane.
Multiplication of either of these vectors by a vector which
is not a function of the time cannot change the frequency, hence
the ordinary laws of complex algebra hold here, and that problem
need not be treated : the method is exactly that of Steinmetz.
However, the multiplication of two class (A) vectors needs to
be considered.
The product of two vectors, as representing a physical quan-
tity, (power), for example:
e, -f ie, + ko
^ -h J''.. + ko
is made u]) of a scalar and a double frequency vector part, — con-
forming to the notation of the X^ector Analysis, 1 shall define
these parts as
(scalar) (e,i) ^ c,i, -\- eJ., A- o o
(vector) [e, i] =
1 J kl
ei e. o\
i\ iz 0 1
The first part is a scalar, the second part a vector, viz.,
fe,i] =kic^i., — eS^)
and jijice it lies in the k-direction, is of double frequency.
so THE ARMOUR ENGINEER [Vol. 4, No. 1
The complete product is
which is exactly the form required for the power. The unit
vector, k, means that its coefficient is wattless, that is, of double
frecjuency.
The meaning of the scalar product, (e,i), is
eizos (e,i)
and of the vector product, [e,i],
e i sin (e,i) k
as is shown in treatises on Vector Analysis. It is also true that
[e,i]=-[i,e]
which perhaps suggested Steinmetz's convention'''
y-i=-i-y
for the representation of power.
The adoption of the double frequency axis, k, permits a real
interpretation of power, in the sense that it may be represented
'graphically as a point in space (instead of in a plane). For,
denoting P by
P' + kP'',
if w^e lay off P' along the real axis, and F'' along the k-axis, a
point is uniquely determined in the (l,k) plane, and this point
represents the power in the circuit, both real and wattless.
The apparent power in the circuit is found from
R^-=(e,iy-+ [e,i]^
and the power factor,
cos« = ^ = ^.
R e ' t
The condition for unity power factor is
^=1-' -1=0.
R\tl t2\
and for zero power factor, D = I.
D is the "inductance factor" of the circuit.
♦Alternating Current Phenomena, Third Edition, page 151,
Jan., 1912] NICHOLS: SYMBOLIC REPRESEXTATIOX 51
The (jcncral wave in the alternating- current circuit is
e = (Zj sin pt + a., sin 2/)/ + . . . .
+ ^1 cos pt + h., cos 2pt -{- . . .
which may he represented syiiiboIicaHy as
e = (a, + j\b,) 4- (a, + y,6J + . . . .
or i= (V, +y\fl',) + (r, +yX) + ....
where the subscripts of the y's show that they refer to different
frequencies. If the a's, etc., represent effective values, the true
power in the circuit is
P' = a,c,-j-a,c,^ ....
. . ^b^d,^b,cL^ ....
= 2(enin)
provided we make the jiatural extension of the diagram to n
dimensions and define the scalar product accordingly.
Similarly, for the circulating wattless power
P" = 2
1 jn kni
On bn 0 | = 2[enin]
^n dn 0
This last is a function of k„ and hence the components of the
circulating power cannot be added algebraically. This must, of
course, be true, for the current and e.m.f. are assumed complex.
The power factor is defined in exactly the same way as
before. Thus the expressions for power are perfectly symmet-
rical in the extended case.
It has been shown that by making the diagram in three in-
stead of two dimensions there is a rational interpretation of power,
both real and circulating, as lengths on the diagram, and that the
already fully developed machinery of A'^ector Analysis gives a
simple solution without the necessity for awkward definitions,
concerning the operations with / for different kinds of products.
THE EVOLUTION OF THE SKYSCRAPER.
BY A. N. REBORI.*
The builder of fifty years ago had very little reason to believe
that the "cast-iron front" which at that time was considered "the
last word" in building construction would eventually prove to be
the beginning of a building activity that today surpasses anything
the world has ever known. It seems incredible, as we gaze along
Michigan Avenue or up and down Broadway that the gigantic
buildings, towering above us from all sides, have practically all
been erected within the past twenty years. However, progress in
architecture does not consist in the mere multiplication of build-
ings, but rather in the real artistic achievement. Admitting that
the aesthetic side of the tall building has been sadly neglected,
that is to say. slighted in comparison with the great amount of
attention paid to the more purely commercial and constructive
elements, still, there are a number of "skyscrapers" that today
stand out in all their bigness, and demand attention and praise
as logical and artistic solutions of a purely American problem ;
and further, the number is gradually becoming greater with the
general advance of knowledge in architectural design and composi-
tion, so that with continued courage on the part of the architect,
endowed with a stiffened backbone, the time is near at hand when
the wise and sagacious owner will be made to realize that in a
building which cries aloud for attention and consideration, which
invites criticism because of its vast bulk and cost, the artistic im-
pression, the form, the outer aspect, are of supreme public import-
ance, and should be and must be given careful consideration and
attention.
Beauty of line and justness of proportion are not arrived at
overnight, but are the result of continued study combined with in-
ventive imagination, scientific knowledge, and artistic ability.
"Nothing can be more depressing than the undertaking to do some-
thing new, by a man who is unaware of what has already been
done, and who has not learned how it is done."
But let us go back a generation or so. and we will find "iron
construction" in its infancy, at the time of the erection of the first
cast-iron front in New York, 1848, when the rising steeples of
the city's churches stood out alone, conspicuously towering above
tJie even building line. Elaborate renaissance fronts, which were
*Associate Professor of Architecture, Armour Institute of Technology.
Jan., 1912] REBORI: SKYSCRAPERS 53
being built of stone and chiefly of marble, were reproduced in
metal, cast in sections, set in place piece by piece, and finally
painted in imitation of masonry. As an artistic expression of a
metallic building, the cast-iron front, which no doubt to our pre-
decessors appeared beautiful, is to us an irrelevant, and unbecom-
ingly cheap, copy of the real thing.
It had its commercial advantages, however, and many promi-
nent merchants of the day were not slow in recognizing the value
of a front that by the nature of its material permitted lighter plan
supports, and at the same time afforded a maximum amount of
window space and light area. The A. T. Stewart Department
Store, erected in 1860, is about the acme of that period ; the build-
ing occupied the entire block bounded by Broadway, Fourth Ave-
nue, Ninth and Tenth Streets, New York, and is still "doing busi-
ness at the old stand" as part of the Wanamaker store. This is
not the first period of metal, however, for previous to this time
cast-iron columns had been in use, and occasionally a few curi-
ously-shaped European rolled-iron sections, principally channelled
or I-beams, found their way at rare intervals into our buildings.
But it was not until 1860 that the first I-beams were rolled in this
country by the Phoenix Iron Company of Pennsylvania and the
Peter Cooper Mills. Trenton, New Jersey, who succeeded in pro-
ducing a seven-inch I-beam, which, used as a lintel, increased the
span of opening. With the rising demand for commercial build-
ings offering greater window area, the development of the germ
of the future steel industry began. This industry not only meets
and supplies the demand of the United States today, but its ship-
ments reach the four corners of the world : to realize better the
leaping strides and tremendous growth of this American institu-
tion, (the steel industry), it is well to remember that at the time
the first I-beams were rolled in this country more iron was im-
ported than produced.
The short period of business inactivity from 1861 to 1866
came to an end with the re-establishment of social relations and
brought with it an increased prosperity. The growth of business,
together with the centralization of interests, so greatly enhanced
the values of favorably located lots that owners were called upon
and fairly forced to build skyward, in order to get adequate re-
turns on their investments ; but this was impracticable, for ten-
ants would not mount stairs above four, or at the most, five stories.
It was here that the elevator showed the way and taught men
to build higher and higher. — for without the elevator a high build-
ing is impracticable.
The era of the skyscraper began with the year 1870, when the
more daring builders went as high as six, then seven and eight.
54 THE ARMOUR ENGINEER [Vol. 4, No. 1
and even nine, stories, and the climax seemed reached. A few
years, and the lesson was learned that such buildings could not
be controlled in the case of fire, and were, to put it mildly, a men-
ace to public safety ; hence, the law requiring them to be fireproof
brought about the first great step in the development of iron con-
struction, and the use of fire-resisting material.
The old New York Post-Office, erected in 1870, is one of the
earliest examples of the use of hollow-tile, flat arches, between
iron floor beams.
With the erection of buildings above the average six-story
limit came the necessity for a safer construction. — one that would
afford fire protection. Following the great Chicago fire in 1871,
a new impulse was given to fireproof construction and the "hollow-
tile beam-covered floor system" came into general use all over the
country. Wooden beams in floors, stairs, elevator enclosures,
and in fact every constructive or exposed part, had to be replaced
by iron, not only to prevent the decay and burning of the wood,
but because the fireproof construction in partitions and floors
added so greatly to the weight.
Owing to the continually increasing valuation of property,
low-storied office buiidings occupying favorably-located lots no
longer yielded sufficient income for the owner. More room was
needed and prices continued to rise ; therefore, buildings had to
go higher. — but here arose a new pre blem, for in the self-carrying
method of building construction the higher the brick or stone wall
the thicker it must be in its lower part, so that brick walls at the
base of a ten-story building became so enormously thick that their
cost was very great, their weight excessive for poor foundations,
and above all the valuable ground occupied by them was a great
loss to the owner. It became necessary, therefore, to make the
walls thinner, and iron construction was resorted to, culminating
in what are now so commonly known as "skeleton constructed"
buildings.
At first, attempts were made to build thinner walls by stiff-
ening them with iron columns at intervals, but this method did
not prove very successful, as the wall joints would shrink while
the columns remained unchanged. Then columns were introduced
to remove the entire weight of floors from the walls, — a much
better device.
In the Home Insurance Building, erected in 1884, probably
the pioneer of its class, various efforts were made to economize
space by the use of both iron-constructed walls and solid masonry
walls. The front walls, facing on La Salle and Adams Streets,
were built of solid masonry up to the second-story level, four and
a half feet thick at the base; from this level to the roof cast-iron
Guaranty Building:, Buffalo.
Jan., 1912] REBORI: SKYSCRAPERS 55
columns connected with rolled beams at floor intervals, were im-
bedded in the walls, the columns carrying the entire floor load
above the second story. The rear walls, excepting a portion of
the light-court, were of uniform thickness throughout, twenty-one
inches, with cast-iron columns fifteen inches at the base, the col-
umns gradually diminishing from floor to floor to eight inches at
the top. The light-court wall, directly behind the elevator, was
built almost entirely of iron and was largely filled with glass, the
solid parts between the columns being only about six inches thick
over all. The side walls were of solid brick. The two upper
stories, the tenth and eleventh, were added five years later. To
show how rapid has been the progress in construction it is but
necessary to state that in the building of the extra two stories,
added in 1890, such a change had taken place that all of the walls
were built with skeleton wrought-iron construction with a sprink-
ling of steel beams, which was probably one of the first consign-
ments from the Carnegie Mills at Pittsburg.
In the Tacoma building, erected in 1887, the exterior walls
were carried independently from floor to floor on the wrought-iron
floor beams, which, in turn, were supported by cast-iron columns
resting on the foundations — which was a marked improvement
in building construction. Cast-iron columns were probably used
on account of the initial saving of about $10,000 over the extra
cost had wrought-iron columns been substituted.
About three years later came the Lancaster Insurance Com-
pany Building, (New York), in which all of the walls were built
with skeleton wrought-iron construction, so that, although the
building rises ten full stories above the ground, the brick side walls
are only twelve inches thick throughout. The skeleton construc-
tion, in which the entire weight of the walls and floors is borne
and transmitted to the foundations by a framework of metallic
posts and beams, reached its full state of evolution about 1890,
after a gradual but remarkable development in building construc-
tion.
Wrought-iron, closely followed by wrought-steel construc-
tion, proved so superior in its economy of space and rigidity, that
cast-iron was eliminated from any further consideration, and with
the assured success of steel, which up to this time had been mostly
experimental, the more pretentious buildings rose to the height of
fifteen and even sixteen stories. With the increased height, and
enormous weight of the superstructure of the tall building a new
difficulty arose, which had to be conquered before a greater alti-
tude could 'be made possible. Piling was used for structures of
moderate height. As to grillage foundations, they were out of
56 THE ARMOUR ENGINEER [Vol. 4, No. 1
the question as far as the skyscraper was concerned as it was not
possible to get .spread enough within the hhnited city lot-Hnes. The
problem really was to carry the foundations of very high buildings
down to bed rock. That was accompli.shed by the "pneumatic-
caisson system," introduced for the first time in 1894 in the case
of the Manhattan Life Building (New York). This method,
however, had long been used in the founding of piers and bridges
all over the world. The pneumatic-caisson system of deep foun-
dations was the third step in the development of the modern tall
building, and ranks scarcely second to the elevator or steel-skele-
ton construction, the other two important innovations, without
which the skyscraper would not exist. The Manhattan Life
Building led the way. and by its .success settled the que.-^tion as to
the proper way of treating similar problems. — for prior to the in-
troduction of the pneumatic-cais.son system, the foundations were
the weak spot in a tall building. We move so very fast in the way
of commercial building that we are apt to forget that the Manhat-
tan Life Building was a ])ioneer. It was one of the first examples
of the possibilities of altitude afiforded by the steel frame con-
struction combined with the pneumatic-caisson foundations
brought into use less than a score of years ago.
With the practicable limit of height no longer an open ques-
tion, and in the absence of any legal restrictions, the New York
skyscrapers began to rise higher and higher toward the sky in a
mad race for supremacy. New York, with its financial district
situated with rivers on either side forbidding lateral expansion,
was forced to find room aloft for the vast interests demanding
office space. In Chicago, however, with the plan of the city per-
mitting expansion in three directions, — North, South and West, —
there is no particular reason why tall buildings should rise up in
the clouds, and further, the building ordinance, which until re-
cently restricted the height of buildings to 260 feet, has once more
been revised, reducing the maximum height to 200 feet, which is
the present altitude limit. No doubt the nature of the subsoil,
which in many cases is not conducive to great height, played some
part in the passing of an ordinance governing the range of build-
ings.
In New York alone it is approximately estimated that 370
skyscrapers over ten stories in height have been erected since
1890. The tallest have 30, 32, 34, 39, 47, 52 and 55 stories ; the
highest are 750, 700, 612, 540 and 486 feet, with a score over 300
feet. These structures, built of "steel-skeleton," protected and
enclosed with imperishable brick, with partitions and floors fire-
proofed with hollow tiles, with window frames and fittings of
West Street Building, New York.
Jan., 1912] REBORI: SKYSCRAPERS
metal, with wire-glass windows, are the safest in the world. They
rest on bed rock, pneumatic-caissons being sunk to the required
depth, 110 to 130 feet in the case of the Woolworth Building, "the
tallest building in the world," now rising rapidly above the street
level to its ultimate height of 750 feet. In the foundations of the
Municipal Building, Xew York, now under construction, over one
hundred pneumatic-caissons were sunk to bed rock, in some places
260 feet below the street and 239 feet below the water level, de-
clared by engineers to be the most difficult foundation ever con-
structed. With the vast strides we have made in scientific engin-
eering, with the present methods of construction at our disposal,
practically nothing is impossible in the way of building achieve-
ment, and today there is simply no limit to the height to which
a building may be safely carried. Where we will stop there is
no telling. The limit seems reached, and the reasonable height ex-
aggerated, still we hear of plans being drawn for a 100-story
building rising 1.260 feet al)ove the sidewalk.
In thus briefly outlining the causes for our recent remarkably
rapid progress in building construction very little or practically
nothing has been said concerning the external treatment or the
aesthetic side of the skyscraper. Now, let us go back to the early
stages once more, and see what real progress we have made toward
a logical architectural expression of the problem involved. To
begin with, the seven-story building, with which the elevator-
building began, or even the ten-story building, with which the
elevator-building culminated, so long as it was built with real
walls, did not bring about an architectural revolution. It was still
possible to follow the analogy of the three-story or of the five-
story building, by making the architectural stories multiples of the
actual stories ; but when the actual stories grew into their teens
and the solid masonry walls were replaced by the skeleton con-
struction, this treatment became no longer feasible.
As there was no further need of self-carrying walls there
was no longer any reason whatever for covering the structural
cage with irrelevant masonry, in an effort to imitate stone. Still
it seemed only natural that the architect, attempting for the first
time to design a tall building, should turn to his fountain of archi-
tectural knowledge and there rake out old "motives," and pro-
ceed to follow the tradition of the stone architecture of the period.
However, the laws of Vignola were not drawn to solve such prob-
lems as those with which the modern architect starts out to illus-
trate them. Surely, the difficulties are not lessened when classic
detail -'s employed, for the mouldings and ornament increase with
58 THE ARMOUR ENGINEER [Vol. 4, No. 1
diameters. At first very little attention was paid to the entirely
new element of design that was let in by the sudden enlargement
of the vertical dimension. After a short period of the "Roman-
esque Revival," which was greatly influenced by the works of the
late H. H. Richardson, the Classic and the French Renaissance
in their various forms became the prevailing influence in the treat-
ment of the facade. Classic columns, pilasters, cornices, and de-
tails were applied in their entirety to the fronts of the tall build-
ings. The designer of the day seems to have taken a special de-
light in disguising the height of his building by the introduction of
a monotony of horizontal lines, brought about by the use of cor-
nices, and in some instances full superimposed orders in groups
of two or three stories. (The Mail and Express Building and the
St. Paul Building, New York, are good examples for illustration
of this point.) After a good deal of experimentation a simple
solution of the new problem was found in one separate treatment
of the bottom and top, and a uniform treatment of the shaft, no
matter of how many stories it might happen to consist. It was
in the Union Trust Building on Broadway that this solution was
first reached and at once commended itself to most designers of
tall buildings in the East, who had not attended to what the archi-
tects of Chicago had been doing. Louis Sullivan was the first
architect to attempt to solve the problem in high design. Almost
from the start he has frankly expressed the vertical elements and
given to high building a logical, as well as a genuinely artistic,
expression. The problem, as he understood it, was to protect
a steel frame, provide all the necessary light in a building devoted
to strictly commercial purposes, and to let the building tell its
own story as agreeably as it might. The Condict Building, New
York City, the Guaranty Building, Buffalo, and the Wainwright
Building, St. Louis, are the most conspicuous examples of his most
personal and thoroughly intelligent efforts. Although he occa-
sionally failed to strictly adhere to his own principle, "form should
follow function," he has shown the way for a further develop-
ment of a characteristic American architecture. It seems a pity
that some of Mr. Sullivan's later and more artistic treatments of
the skyscraper problem do not exist in Chicago, the home of the
famous architect.
"The columnar treatment," with its base, shaft, and capital,
as a motive for the exterior treatment of the skyscraper, is most
logically and artistically expounded in the "Broadway Chambers"
(New York) with its rusticated stone base, its simple rough, red-
brick shaft, and its capital of vari-colored terra-cotta. The design
as a whole is extremely well handled, with a certain simplicity,
Woolworth Building, New Yorls.
Jan., 1912] REBORI: SKYSCRAPERS 59
and a rather pleasing use of external color, which tend to give
movement and variety to the flat elevation. A small scale model
of this building was exhibited at the Paris Exposition of 1900.
where it was awarded a "diploma of honor," being highly praised
by the jury of award for its rigid adherence to conditions, besides
excellence in design.
The Corn Exchange National Bank Building, at the corner
of LaSalle and Adams Streets, (a local example of the base,
shaft, and capital composition), stands firmly set in position with
a solidity which is the result of a frank treatment, regardless of
the encased steel-skeleton frame actually doing the work. Aside
from this non-observance of its structural body, the elevation
is consistently studied throughout from its strong stone base and
banking story up through its large, flat shaft of brick, to its crown-
ing top of more ornate terra-cotta. The base of the exterior walls
projects about eight inches beyond the shaft above, giving a taper-
ing appearance to the building, when viewed from a distance,
which is quite effective. The whole surface is treated in a mono-
tone, no color effects being attempted.
With the atmospheric conditions of our city, inimical to the
external use of color in buildings, this simple method of obtain-
ing pleasing effects has been temporarily abandoned. After the
successful use of enameled terra-cotta, introduced for the first
time in 1894, as an outer covering for the steel frame of the Reli-
ance Building at the southwest corner of Washington and State
Streets, the local demand for white enameled terra-cotta has grad-
ually grown to vast proportions. The chief virtue of this material
lies "in the fact that it can be easily cleaned by the simple process
of washing. The exterior use of white enameled terra-cotta has
created a sort of bathroom renaissance architecture, which looks
much more like a skeleton covered with a thin skin than a real
body properly clothed in flesh.
There is no reason for going to extremes ; architecture is
something more than just the plain frank expression of truth.
The Ingalls Building, built on the northwest corner of Fourth
and Vine Streets, Cincinnati, Ohio, deserves mention, principally
because of its being the first concrete skyscraper. It was begun
in the fall of 1902, having required in its erection a little longer
time than the standard "steel-cage" type of the same size. The
building occupies the entire area of a corner lot 50 by 100 feet,
and is sixteen stories, rising to the height of 200 feet above the
sidewalk. The structure from the bottom of the foundation is
235 feet, entirely concrete steel. In reality it is a concrete box
of eir?ht-inch walls, with concrete floors and roof, concrete beams,
60 THE ARMOUR ENGINEER [Vol. 4, No. 1
concrete columns, concrete stairs, the whole entirely devoid of the
usual I-beams, angle-irons, plates, rivets, and bolts. It consists
merely of bars imbedded in concrete, with the ends interlaced,
making a complete concrete monolith of the entire building; all of
which is most reasonable and expressive until we arrive at the ex-
terior, which we find is covered with a veneer of from four to six
inches of white marble for the lower three stories, glazed-brick
for the next eleven, and glazed white terra-cotta for the top story
and cornice ; immediately losing its identity and causing the build-
ing to look for all the world like an ordinary modern steel building
with a marble, brick and terra-cotta covering. Inasmuch as a
concrete building is not built up like masonry, and at once becomes
a monolithic structure, every particle of which is doing structural
duty, it seems illogical, to say the least, to attempt to hide the
truth of its real being by the mere sham of covering up its face
with brick and stone. ( Perhaps the same could be said of a steel
structure, — except that in the case of steel the law demands the
external covering of fire resisting material.)
Until the architectural forms, mouldings, and decorations
are incorporated with the moulds used for the structural work,
and the surface of the exposed concrete is treated either in a di-
rect or a more homogeneous manner, very little can be expected
from this use of concrete. Something might be done by applying
several thin-finish coats in different colors, on the exposed surface,
to be treated architecturally. The design could then be tooled
through the outer layer, exposing the sub-layers in contrasts of
light and shade, very much like the early Italian scrafitto work.
Color might be used in various tones, in fact anything that
would tend to give an expressive treatment and an appropriate
decoration to the material used. Then perhaps we may hope
for a truly rational architecture, one in which there is no sham, no
deception ; a solid without joints, — every member incorporated.
Most of the office buildings erected by the "rein forced-con-
crete method" have been faced with brick and stone. The few
examples that have attempted to depend solely upon concrete have
stuck pretty close to the precedent of masonry, not attempting
a more direct expression of the individuality of concrete, thaii
the avoidance of an excessive pronouncement of stone.
Inspiration, "the act of exercising an elevating influence upon
the intellect or emotion," as applied to architecture, cannot be of
substantial value unless derived from the actual structure. To the
man of ability, inspiration is usually the sudden desire to get
down to work. The West Street Building is without doubt the re-
sult of applied inspiration ; it is also the result of a great deal of
ii«"i:
II II II n ij tt
ii II II •■ II
II Ii ii 11 II
[III
"l[||iiisii;i
:(;r
Monroe Building:, Cliicago.
Jan., 1912] REBORI: SKYSCRAPERS
preliminary study. The architect clearly endeavored to permit the
structure to design itself, confining his own role as much as possi-
ble and as long as possible to making the structural features as
good looking as lay within his power. The result is obvious.
The steel frame is covered with fire-resisting material, and the
open spaces are filled with glass where glass is required, and the
ornamentation is confined to such expression as rightfully can be
imparted to the material used. The long, vertical shafts are grace-
fully terminated above the rich cornice by a series of pointed terra-
cotta dormers. The corner pavilions are strengthened, culminat-
ing in the slender domed Belvideres, which in themselves are ad-
mirably handled. The copper roof Is highly colored and the care-
fully-placed color accents add considerable charm to the design.
The building as a whole is well studied and efifective In compo-
sition, scarcely surpassed In its own line by anything that has since
been done.
The same architect, Cass Gilbert, Is the architect for the
Woolworth Building, Broadway (Barclay Street to Park Place).
This stupendous structure, begun in 1910. to be comoleted by the
fall of 1912, covers a plat 152 feet by 197 feet, and will rise, when
finished, to a height of 750 feet from the sidewalk, fifty-five
stories above the ground. The main building contains twenty-nine
stories, with the tower, 86 feet by 84 feet, rising another twenty-
six stories. — the "tallest building In the world." The design as
presented is the result of over a score of complete preliminary
studies, all of which were carefully rendered and carried to the
limit. The caisson foundations are already in place and the steel-
skeleton is well under wav. while the detail for the top stories is
still undergoing considerable study toward possible improvement.
In its artistic conception, the Woolworth Building ranks with the
greatest monuments of the world.
The latest addition to the Lake Front sky-line Is the Monroe
Building, at the southwest corner of ^lichlgan Boulevard and
Monroe. That it is a simple and direct architectural interpreta-
tion of the structural requirements Is at once evident. The photo-
graph of this building, taken during construction, reveals the
upper portion of the steel frame still exposed, and the covered
portion with its finished terra-cotta shell, fairly climbing up and
around the steel columns and across the floor girders, droppincj
into place as naturally as the bark clings to the tree. The huge
gable roof is an innovation from the regulation-type flat roof, so
commonly used, and alone for this reason invites criticism. It is
quite evident that the picturesciue treatment of the roof on the
Monroe Building is rather an attempt to recall the silhouette of
62 THE ARMOUR ENGINEER [Vol. 4. No. 1
the University Club across the street, than a logical expression
of the functionary duties of the building itself. However, it is
best to withhold judgment until the proper completion of the
building and then perhaps it may be wise to wait a while longer
and give the structure a chance to speak for itself, before offering
a hasty criticism.
The walls of the two lower stories of the building, acting as a
sort of base for the long vertical lines of the superstructure, are
faced with polished pink granite. The entire superstructure is
encased with "standard-finish" terra-cotta of two shades in alter-
nating courses or stripes, very much similar in effect, although in
a much minor key, to the marble interior walls of a Sienese Cath-
edral. The thin pilasters attached to the granite base are much
too delicate and entirely inadequate as continuations of the solid
vertical piers above. By strengthening the end pavilions above
the second story, (in themselves well studied and effective), and by
frankly ignoring this emphasis in the treatment of the bays of
the two lower stories, two distinct elements of composition have
been introduced whose relation, one to the other, marks the weak-
est part of the general design. As a whole, the building has been
carefully studied, with an excellent arrangement of large win-
dows, raised high above the ceiling, broad and low and shaped as
they ought to be for utilitarian purposes. The Monroe Building
will surely act as a stimulus for greater efforts along the same
line, and for that reason, regardless of its more purely arcitectural
effect, the designers deserve a great deal of credit for what they
have done.
In the modern commercial building the problem of the interior
is chiefly one of construction. The entrance, the lobby, the ele-
vator, the hall, the corridors, and an occasional banking room, are
legitimate places for the display of the architect's personal taste.
The rest of the plan is usually arranged with each floor as one
great loft to be subdivided by light, interchangeable, and easily-
moved partitions to suit the tenant's wishes.
There are a number of eminent architects who know and take
cognizance of the fact that the high building problem is not one
that will solve itself; but it can only be solved by the most pains-
taking care, by the most thorough study of past efforts and fail-
ures, and by a thoroughly artistic meeting of all of the conditions
involved.
Reasonable accjuaintance with the principles of structural
engineering is of prime importance to the architect in the practice
of his profession as a fine art. The day for the architect of ar-
tistic temperament who scoff'ed at and was bored by the engineer
Jan., 1912] REBORI: SKYSCRAPERS 63
and frankly admitted that he knew nothing about engineering,
has gradually passed with the coming of the skyscraper. At the
present day, which in all probability marks the zenith of the iron
age in building construction, the architect and the engineer must
work hand in hand, in order to achieve the best results. The en-
gineer makes accurate and elaborate calculations of difficult prob-
lems in construction, his primary object being to obtain the maxi-
mum strength with the minimum of material, while the architect
pays more attention to the aesthetic side of engineering ; it is for
him the art of designing the structure of his architecture, be it
stone, wood or metal, in a serviceable and highly artistic manner.
The greatest architects in history were also engineers. Not
engineers, however, as the term is understood today, technically.
Surely the great Dome of St. Peter's or the Ancient Pantheon in
Rome, with its span of over one hundred feet, can bear a rigid ex-
amination with regard to their constructive excellence. We should
no doubt find them as worthy of being held up as examples of
emulation for their structural quality as for their more strictly
architectural merit.
The skyscraper problem affords the architect ample oppor-
tunity for presenting the two elements in a just relation to each
other. There is no doubt whatever that the logical architectural
solution of this type of building is to be sought today scientifically
in the proper relation between architecture and engineering.
Modern engineering is indeed a very modern thing. With the
enormous advantages which it has brought to the world, it has
brought this disadvantage, — that for the first time in human his-
tory a broad line has been drawn between scientific construction
and artistic construction, and that the designers of the one class
of construction do not hold themselves responsible — nor does any
one else — ^for the looks of their work; take for example some of
the bridges across the Chicago River. If an engineer builds safely
and cheaply, and in a word scientifically, his work may be as ugly
as it please without any impairment of his professional reputation.
In architecture, the recognition of permanency is one of the
true principles of the art. A front must not only be strong
enough, 'but it must possess an evident reserve of strength, which
is the result of obvious abundance. A building should bear the
impress of solidity, as though it were indeed a growth of the earth
itself, and not of so fragile an appearance that the wind can blow
it away.
From the owner's point of view, the architect is a practical
man, serving his client to the best of his ability, sacrificing no inch
of room anywhere to architectural effect, but employing every
means of utiHzing the area and the altitude. Certainly a sacrifice
64 THE ARMOUR ENGINEER [Vol. 4, No. 1
of any practical requirements to appearances is not only "bad
business," but bad faith and likewise bad architecture. However,
a building may have all the practical requirements demanded of
it, besides being as logical as possible and yet be ugly, when it
might instead be made highly artistic and effective, by reason of
the skill of the architect in his emphasis and his subordination, in
the artistic spacing of his decoration, in the placing and scale of
his detail, in the study given to his design as a whole, based on
function, reason, and logic.
Popular judgment upon buildings, as works of art, is mostly
vitiated by the thoughtless habit of ascribing to the architect his
advantages as merits, and correspondingly, his disadvantages as
faults. Criticism must be kept clear of this confusion. The
problem confronting the architect in any case is to make the most
of the advantages and minimize the disadvantages, and to do these
things with the least sacrifice of the strictly utilitarian purposes of
the structure, and yet to make as expressive, harmonious and beau-
tiful a building as the conditions permit. This can be only accom-
plished by the most earnest and the most conscientious study of
the problem involved. To convert difficulties into o])portunities
should be the aim of every architect worthy of the name.
RECENT DEVELOPMENTS IN WIRELESS TELEGRAPHY.
BY LOUIS COHEN.*
Among the numerous contributions to the progress of wire-
less telegraphy in recent years, there is probably none which ranks
equal in importance to the development of methods for producing
sustained oscillations, and the development of receiving apparatus
which is responsive only to a sustained oscillation of definite fre-
quency : to appreciate its importance we must know something of
the difficulties met with in the actual practice of the art. In the
early days of wireless telegraphy when there were only a few
isolated stations in operation and at considerable distances from
each other, the problem of interference was not serious ; the work-
ers in the art were mainly concerned with the problem of estab-
lishing communication between stations irrespective of efficiency
and reliability. With the rapid multiplication, however, of the
number of stations, and particularly the increase in number of
high power stations, and also with the demand made upon stations
to work continually, the problem of interference assumed great
importance.
It is quite obvious that if there are several stations working
simultaneously and in close proximity to each other, it becomes
impossible to maintain communication between any two stations
unless we have some method for eliminating, or at least reducing
to a minimum, the interference caused by the other stations. In
seeking for a solution of this problem it occurred to several inves-
tigators and inventors to utilize the principle of electrical reson-
ance, which was well understood at that time. Every electrical cir-
cuit containing inductance and capacity has a time period Avith
which it will oscillate if it is disturbed from its electrical equilib-
rium, and it will naturally respond more powerfully to properly
timed electrical impulses of the same period, and in that way make
any wireless station more selective, so as to receive signals from a
certain station only and not from any other. However, to bring the
l)henomenon of electrical resonance into full play two things are
required: the electrical waves acting on receiving antenna must
be more or less sustained, at least ten to twenty oscillations per
train of waves, and the receiviug apparatus must be of such form
as to respond more powerfully to a sustained train of waves than
to an isolated electrical impulse.
"'Class of 1901. Chief, Rese.arch Department, National Electric Signaling Com-
pany, Brant Rock, Mass.
66
THE ARMOUR ENGINEER
[Vol. 4, No. 1
Let us consider first the developments and improvements in
the sending apparatus. The method first adopted by Marconi for
generating- energy in antennae is shown in Fig. 1. The secondary
of an induction coil was connected across a spark gap which was
inserted in the antenna ; the gap was adjusted so as to spark across
when the potential in the secondary of the induction coil built
itself up to its maximum value. The charging energy of the
antenna is £ = J^Cf'- and at N sparks per second the total
energv is
The Armour En<j inter.
Fig. 1. Simple Marconi Sending Apparati
Assuming the following constants :
A^ = 30 sparks per second. C = 0.001 microfarad,
[' = 20,000 volts ; then £ = 6 watts.
The energy per spark is only about 0.2 watt.
When the spark occurred the antenna discharged itself, con-
verting the electrostatic into electromagnetic energy, reversing
itself again, and continuing that way until the entire energy was
dissipated, partly in heat due to ohmic resistance of antenna and
partly in radiation. In such a form of oscillator, the oscillations
are few, say a half dozen or so, and are necessarily highly damped,
the small amount of energy being rapidly frittered away by re-
sistance and in radiation. Owing to the small capacity of the
Jan., 1912J
COHEN: WIRELESS TELEGRAPHY
67
antenna, it is impossible to store up, a large amount of energy,
which makes it impossible to produce feebly damped oscillations.
Furthermore, the electromagnetic waves emitted by sending an-
tenna diminish in intensity with increase in distance, and neglect-
ing absorption, which will be discussed later, the intensity varies
inversely as the distance; therefore the greater the intensity of
the radiated electromagnetic waves, or the more energy available,
the larger the distance that can be reached between stations, and
hence the desirability of developing methods and apparatus which
should make it possible to utihze a large amount of energy.
A very great step in the advancement of the art was accom-
plished by the introduction of the oscillation transformer shown
Highly Dam pi
The Armour Engineer.
I'iS. 3. Types of Oscillations.
in Fig. 3. This may be taken to represent diagrammatically the
arrangement used in every station at the present time. In all
but the smallest stations the induction coil was replaced by an
alternating current generator A and power transformer P . In
some stations the auto-transformer, or as it is sometimes called,
direct coupling, is used in place of the ordinary oscillation trans-
former T, or electromagnetic couphng, — ^the principle, however,
is the same in either case. This arrangement ofifers the advan-
tage that we can use a large condenser in the primary circuit
L G C and therefore store up a considerable amount of energy.
When the condenser is charged up to the sparking potential of
gap G, and a spark occurs, the energy stored up in condenser is
liberated and it sets up oscillations in circuit L G C, which also
68
THE ARMOUR ENGINEER
[Vol. 4, No. 1
induce oscillatory currents in the antenna circuit and thus transfer
part of the energy to the antenna. The energy in primary oscil-
latory circuit will be gradually damped out, partly by the resist-
ance of the primary circuit, and partly by the transference of the
energy to the antenna circuit where it is dissipated by resistance
and radiation. However, this arrangement ofifers the disadvan-
tage that the coupling which is defined by
M
Vuu
must be made very loose. If we increase the coupling and thus
T 777^ Armour Engineer.
Figr. 3. Sending Apparatus with Oscillation Transformer.
hasten the transference of energy from primary to antenna cir-
cuit the energy surges backwards and forwards and produces the
phenomenon of beats. Instead of generating electrical oscil-
lations of one definite frecjuency, we have two oscillations of
diiTferent frequencies, and the difference in frequencies will be
larger the greater the coupling; so that instead of generating
waves of one definite length corresponding to the natural period
of the antenna, we have two waves, one of which is above and
the other below the natural wave-length of the antenna circuit.
This is objectionable because it does not permit very sharp tuning,
for instead of sending out one wave-length, say 1,000 meters, a
pulsating oscillation is sent out which may be looked upon as
Jan., 1912] COHEN: WIRELESS TELEGRAPHY
69
made up of two damped waves, of say 900 meters and 1,100
meters. It is obvious that any receiving station tuned for 'any
wave-length in the region 900 to 1,100 meters will respond to these
oscillations.
To eliminate the difficulty of double frequency, many efforts
were made to devise methods for breaking the primary circuit
very quickly so as to avoid having the energy surge back and forth
from primary to secondary, by having the entire energy trans-
ferred to the antenna circuit during the first one or two oscil-
lations, and allow it to oscillate in antenna circuit alone. This
would permit the use of strong coupling and at the same time
have only one frequency. The first practical solution of this
problem was obtained by replacing the stationary spark-gap by a
rotating spark-gap, which affords a good means for opening the
primary oscillating circuit very quickly. Another form of spark-
gap which offers a good solution to the problem and which is now
being rapidly introduced into commercial practice is the so-called
quenched spark-gap. In this form of gap the spark is made to
occur between parallel copper or silver surfaces separated only
0.01" to 0.0015". The discs are separated by a thin annular ring
of mica or rubber which also serves to shut the sparking spaces
off from the air. In this form of gap the spark is quickly
quenched, and the entire energy is transferred during the first
oscillation. Several gaps are generally used in series, the number
depending on the power and potential of the system.
The choice of size of condenser to be used in primary circuit
depends practically on the energy which is to be used at the send-
ing station. Suppose we desire to utilize 5 kilowatts, and let us
assume that the alternator is of 500 cycles, giving 1,000 sparks
per second ; also assume the voltage of power transformer to be
20,000 volts, then we have
Energy = 'ANCV^
or 5000= ^ X 1000 X C X (20000) =
therefore C = 0.025 microfarad.
Another advantage of using an alternator with a power
transformer in place of an induction coil is that we can obtain a
large number of sparks per second, depending on the number of
cycles of the alternator. In the case of an induction coil the
number of sparks per second was limited by the mechanical make-
and-break, which never exceeded 30 or 40 per second ; with the
alternator, however, we can easily obtain a thousand sparks per
second, and this is the spark frequency which is now being com-
monly used. The spark frequency determines the note which is
heard in the receiving telephone, and the higher this note the
70 THE ARMOUR ENGINEER [Vol. 4, No. 1
more readily can we distinguish it from any irregular noises, par-
ticularly those caused by atmospheric disturbances. The arrange-
ment of circuits shown in Fig. 3, with the use of rotary or
quenched spark-gap, gives feebly damped oscillations, anywhere
betwen 20 and 100 oscillations per train of waves, — which makes
tuning of the receiving antenna possible.
Considerable work has been done in developing methods for
producing undamped oscillations. In Europe the efforts seem
to be directed to the improvement of the arc for this p\irpose,
but so far the glittering promises which have been claimed for it
4. High Fre<iuenoy Alternator for Producing Persistent Oscillations, Gear-
connected to Direct Current Motor. Alternator, 2 K\V, 20,000 r.p.ni., 110
volts, 100,000 cycles per second. Motor, 2,000 r.p.m., 120 volts.
have not been realized. In this country the National Electric
Signaling Company has been develnning the high frequency alter-
nator for the production of persistent oscillations. At the present
time there are several 100,000 cycle alternators in operation, each
having an output of about 2 kilowatts ; one alternator of 50,000
cycles was recently built having an output of 35 kilowatts.
Turning now to a consideration of the receiving apparatus,
we find that the efforts of inventors have been centered on the
improvement of the detector, which forms the most essential
element in the receiving station. It must be observed that the
Jan., 19121 COHEN: WIRELESS TELEGRAPHY 71
receiver must be of suitable form, correspouding to the trans-
mitter. It is evident that in the case of a highly damped radiator
we must have a receiver which is afifected by the first or maxi-
mum oscillation, and this must be inserted in a receiving circuit
which is easily set in oscillation by single, or at most a few, elec-
tro-magnetic impulses ; the coherer answered this purpose satis-
factorily and was the first form of receiver used. It is also mani-
fest that such form of receiver will be affected by any stray elec-
tro-magnetic impulses. If the transmitter is a feebly damped radi-
ator, it will be advantageous to use a receiver which responds
to the cumulative efifect of all the oscillations in a train of waves.
This permits the use of a stifif circuit and hence is not readily
afifected by any stray electromagnetic impulses. In other words,
by using a feebly damped radiator we can bring into full play
the phenomenon of syntony. It is possible to use a very stifif
receiving circuit, large inductance and small capacity, and thus
make it respond only to waves of definite length.
The coherer is called a potential operative device for the
reason that it requires a potential difference of definite intensity
to be impressed on it to make it respond ; so that only the first
oscillation, which has the maximum intensity, is effective. Nearly
all other detectors now commonly used depend for their action
upon the current flowing through them, and are commonly called
current-operative devices. In the latter form of detectors it is
the cumulative effect of all the oscillations in a train of waves
which affect the receiver, and as a consequence the entire energy
is being utilized in the receiver. The coherer has now practically
disappeared and was replaced by one form or another of the cur-
rent-operated receivers, the UKXst common forms of which are
the electrolytic detector, or barretter, the crystal detector, and
the magnetic detector.
Professor Fleming illustrates by a verv apt aiialogy from
optics the difference in the two types of receivers. When we look
through a telescope at the stars we can see a certain number down
to some limiting magnitude. No amount of prolonged gazing
when using the eye as a wave receiver increases the effect pro-
duced by a star just invisible. If, however, we use a photographic
film, the effect on it is cumulative and we can, by a sufficiently
long exposure, obtain impressions of invisible stars in countless
numbers. The photographic film is a wave detector of quite a
different kind as compared to the retina. In the case of the film
it can make up by time what is wanting in intensity in the wave
motion. The coherer-type of receiver corresponds to the retina in
the above illustration, while the currcnt-o]:)erated detector corres-
ponds to the film.
12
THE ARMOUR ENGINEER
[Vol. 4, No. 1
We cannot here go into a theoretical discussion of the prin-
ciple of operation of the various detectors, as there seems lo
exist a considerable difference of opinion about the principles
governing their action. The typical arrangement of receiving
circuits is shown in Fig. 5.
The method of using a feebly-damped radiator and a current-
operated receiver, and thus making it possible to bring into full
play the phenomenon of resonance, was a very great step in the
development of wireless telegraphy. Though it has not entirely
eliminated interference, yet it is possible for stations of different
wave-lengths, say twenty-live per cent difference, to operate
iiiiiii
IIM^^
The Armour Engineer.
Fig. .5. Typical Arrangement of Receiving Circuits.
simultaneously without interfering with each other. Many at-
tempts were made to reduce interference still further, and a
device which was developed by the National Electric Signaling
Company, and which gives very satisfactory results, is the Fes-
senden interference preventer. In this device, shown m Fig. 6,
the current generated in the receiving antenna divides at the point
A^ part going through the branch F and part through the branch
D. The branch D is tuned to the wave-length which we desire
to receive, while the branch F is slightly out of tune, and there-
fore the signals to be received will pass almost entirely through
branch D and very little through branch F. The current due to
any other signal of different wave-length will divide itself prac-
Jan., 1912]
COHEN: WIRELESS TELEGRAPHY
n
tically evenly between the. two branches. The secondaries G and
H are so arranged that the induced currents oppose each other,
and as a consequence the effects of any stray impulses or signals
from interfering stations will neutralize each other because they
are of the same intensity in both branches of the secondary cir-
cuit.
Atmospheric or electrostatic discharges very frequently cause
considerable annoyance to the operator, and in tropical countries,
where the atmospheric discharges are very heavy and continuous,
1 1 1 1 1 1 I
The Armour Engineer.
Fig:. G. The Fessenden Interference Preventer.
this matter causes considerable trouble, making it extremely dif-
ficult, if not impossible, to receive weak signals. This is to some
extent overcome by using a high spark-frequency, the most com-
mon now in use being 1,000 sparks per second. The atmospheric
disturbances produce a low rumbling noise in the telephone ; by
using a high note it is more easily distinguishable amidst the noise
caused by atmospheric disturbances. The 1,000-spark note has
another advantage. — it was found that the ear is most sensitive to
this note, and weaker signals can therefore be detected.
74 THE ARMOUR EXGTNEER [Vol. 4, No. 1
At the present time a new method for receiving signals is
heing developed, at the .Xational Electric Signaling Company,
which dififers radicall}- in ])rinciple from every receiver now in
operation. One of the advantages is that it eliminates inter-
ference and atmospheric disturbances entirely. Owing, however,
to the patent situation, I am not at liberty to fliscuss it here.
While great credit must be given to the men who have as-
sisted and promoted the practical development of the art, we must
r.ot neglect to give due credit to the scientific investigators who
have contributed very materially to the advancement of the same.
A great deal of scientific investigation has been carried on in
recent years on problems pertaining directly or indirectly to wire-
less telegraphy, and the results have led to new discoveries and
improvements. In this as in any other art. scientific investiga-
tion must be carried on simultaneously with the practical develop-
ment to obtain the best results. Every new development brings
into existence problems for the scientific investigator which re-
quire an explanation of the principle involved, the determination
of constants, or the development of methods of measurements.
On the other hand, the results of scientific research frequently lead
to some practical improvements or developments. In the study
of wireless telegraphy a large number of problems have arisen
which have required the combined skill of the mathematician and
the physicist to unravel and interpret ; it is sufficient to mention
only a few of the important problems, such as the radiation con-
stants of antenna, the theory of coupled circuits, the resonance
transformer, the influence of frequency on resistance of w^ires,
plates, coils, etc. We must also consider the fact that we are
dealing with very high frequencies, and the methods of measure-
ment which were suitable for low frequencies will not be adapt-
able to this work. — hence new methods of measurement must be
devised.
Before closing I wish to call attention to a very interesting
series of experiments which have recently been carried on by the
United States Government experts, in conjunction with the
engineers of the National Electric Signaling Company, to deter-
mine the absorption constant. It has been known for some time
that the intensity of signals received over long distances dififer
from day-time to night-time, being always much weaker in day-
time. The explanation offered for this phenomenon is that the
rays of the sun ionize the air and make it more conducting; hence
the electromagnetic waves sufifer greater absorption or damping
in their transit. The object of the experiments was to determine
the numerical value of the absorption constant. The experiments
Jan., 19121 COHEN: WIRELESS TELEGRAPHY 75
were conducted between Brant Rock, Alassachusetts, where the
National Electric Signaling Company has a 100-kilowatt station,
and two Government Scout Cruisers, each of which had a 10-
kilowatt station. The Scout Cruisers traveled out a distance of
3,000 miles from Brant Rock and observations on the intensity of
signals were taken at very frequent intervals. From the large
number of observations the following empirical formula was
obtained :
where /„ is received current, K is a constant depending on the
intensity of current at sending station, d is the distance in kilome-
ters, A is the wave-length in kilometers, and a is the absorption
constant.
The results of these experiments are of considerable impor-
tance, inasmuch as it makes it possible now to determine with
some degree of accuracy the probable distance that can be covered
with a given output.
Considerable work has also been done, and much has been
accomplished, in the development of wireless telephony, but space
will not permit us to enter into a discussion of this subject. We
may remark, however, that wireless telephony offers a very at-
tractive field for investigation and research.
It is a matter of great encouragement to note the remarkable
progress, in the art of transmitting intelligence by means of ether
waves, that has been accomplished in a comparatively short time.
At the present time there are about two thousand stations in oper-
ation in every part of the world, contributing very materially to
the safety of life and property at sea. Every passenger liner
is now equipped with a wireless station, making it possible to be
in touch with land during its entire travel across the ocean. Wire-
less telegraphy has developed into an implement of immense im-
portance in naval warfare, so that every important navy in the
world has adopted it as an indispensable means of communication.
There is now no doubt that wireless telegraphy offers a reliable
means of communication, and while we may feel some pride in
the achievements that have already been made, it is but necessary
to glance around to note that there are many unsolved problems in
connection with this subject which offer a very fertile field for
further research and investigation.
WIRE ROD ROLLING.
BY. J. S. BANTA, M. E.*
The function of a Rod Rolling ]\Iill is to produce small-sized
rods from which wire is drawn. The raw product from which the
rods are rolled is a steel billet weighing from 150 pounds to 210
pounds, and usually four inches square, or a copper ingot weigh-
ing from 225 pounds to 275 pounds. The smallest rod rolled to-
day is size 5, or about .207 inches in diameter. Rods as small
as number 8 size or .162 inches diameter have been rolled, but
this was found not to be economical ; i. e., it is cheaper to reduce
size below number 5 by drawing cold than by rolling hot. Rods
as large as 1-1/16 inches diameter are sometimes rolled and coiled
on reels to be later drawn to coarse wire or drawn bars ; rods
coarser than this are not coiled, as a rule.
Generally speaking, the process is as follows : Steel or cop-
per billets are heated to required temperature, rolled in a number
of passes, and coiled on high speed reels as fast as finished in the
mill. To accomplish this end economically requires a large outlay
of money in buildings, heating furnaces, roughing and finishing
mills, machinery to handle cold and hot steel, automatic reels,
engines and boilers, and finally means to transfer product to
the wire mill, or to load for shipment.
In a modern high-tonnage plant the steel is usually loaded
direct from cars to furnace charging machines, or hydraulic push-
ers as they are called. This is usually accomplished by an oyer-
head electric trayeling crane with an electric magnet hanging on
hook. The crane is arranged to span the hydraulic chargers and
standard gauge gondola cars in which steel is received. Ten or
twelve billets are arranged in car side by side by laborers, then
the crane operator runs the magnet over the pile, drops down to
the billets, turns the current on the m'agnet while dropping, and
hoists the billets clear of the car, then makes a run for the hy-
draulic charging machine requiring billets, after which he repeats
the operation. Such a crane will keep heating furnaces supplied
with steel, handling from 350 to 450 gross tons per turn, or make
350 to 450 trips per turn of eleven hours. The magnet usually
used is about forty-two inches diameter and weighs about 2,600
*Class of 10(13. Works Eii!j;iiieer, Americfin Steel & Wire Cninpan.v, Wauke,L;nii,
Illiuois.
Jan., 1912J BANTA: WIRE ROD ROLLING 11
pounds, and is known as a pig magnet. Such a magnet is capable
of lifting live to eight tons in one solid chunk, but only about one
gross ton when divided into billets weighing 200 to 210 pounds.
It is impractical to lift more than one layer of billets at a time.
The billets are usually charged into heating furnaces by
stationary hydraulic pushers, which consist of a long stroke hy-
draulic cylinder operated with water at 140 to 1,000 pounds pres-
sure, or more, depending on diameter of cylinder used. The total
pressure required in pusher on a modern furnace ranges from
22,000 to 30,000 pounds. These pushers merely push a long layer
of billets on skid pipes into furnace.
The billets are heated to a temperature of 2,000 to 2,200 de-
grees Fahrenheit in long gas-fired heating furnaces arranged for
two rows of billets from three feet to four feet long, or one row
of double-length billets. The double-length billets are cut in two
before rolling. The furnace is built of common brick masonry
lined with nine-inch fire brick and all bound together by cast iron
plates, steel buckstays and rods. The billets are pushed in on skid
pipes by hydraulic chargers or pushers mentioned above, all of
which are controlled by one man from a pulpit located so as to
afiford operator full view of pushers, charging end of heating
furnaces, roughing mill and portion of hot-billet conveyor next
to mill. The billets are discharged from far end of heating fur-
nace directly on a hot-billet conveyer or carrier by pulpit man
referred to above. Hither natural or producer gas is used to heat
billets. If producer gas is used, it is generated from bituminous
coal and passes through suitable flues to discharge end of heating
furnace where it strikes the air blast and is ignited, from this
point the flue gradually enlarges to take care of expansion of gas
until it enters the combustion chamber or hot zone. The products
of combustion and burning gases after passing through the com-
bustion chamber pass over and under the billets to the charging
end and down to underground flue to stack at a tempera-
ture of from 1.000 to 1.200 degrees Fahrenheit. In some plants
this heat is utilized to heat boiler feed water or air blast, some-
times both. When air blast is preheated it usually passes through
checker work or pipes located in flue between furnace and stack,
or through air flues located in masonry below furnace, or through
space between furnace roof and a secondary arch.
The gas producers usually used are water-sealed and arranged
with automatic machinery to feed and distribute coal. Coal re-
cjuired to heat a gross ton of billets varies from 200 to 300 pounds.
depending on grade of coal and local conditions. The coal is
crushed to about one inch.
Jan., 19121 BANTA: WIRE ROD ROLLING 79
The best type of hot-billet conveyor consists of a series of
rolls about twelve inches in diameter and driven by motor through
a long shaft and bevel gears. The billets are transferred to mills
at 100 to 130 feet per minute, and are slowed down to less than
half that si)eed just before they reach the mill.
A modern roughing mill consists of six to eight continuous
stands of rolls usually driven by pinions, cross-shafts and bevel-
gearing from main-shaft directly connected to engine-shaft by flex-
ible coupling. Flexible couplings are also provided between rolls
and pinions to provide for variation in diameter of rolls, Cjuick
changes of rolls and replacement of broken parts, also for a weak
place to break in case something unusual happens in the mill.
The speeds of rolls increase from entering end to discharge end
so as to provide for the elongation of the steel billet as it is re-
duced in size, the entering billet being four inches square and the
discharging billet from 1 inch to iVs inches or from % square inch
to 1 square inch in sectional area. To obtain this variation in speed
bevel gears of different ratios are provided to drive each cross shaft
from main shaft. Theoretically the product of the sectional area
and speed of steel passing through each roll should be constant,
but in order to run mill successfully this has to be humored a
little. If this product varies to any extent on one pair of rolls
from that of the other rolls, there will be danger of steel buckling
or cobbling one side of the roll and stretching or drawing out on
the other side, in the first case causing loss in scrap and danger to
mill men, and in the second case danger of producing steel vary-
ing in size. It is good practice to put a little stretch in steel be-
tween rolls; further, surface speed of rolls should not be taken
as speed of steel but should be figured from the pitch diameter of
roll, which varies with the shape of pass and the extent to which
pass is filled out by the steel. A pass can be under-filled, but
should never be over-filled, as this will produce fins and slivers.
both very objectionable wdien drawing wire rod down to wire, —
also producing poor wire and scrap in the wire mill.
As the hot billet reaches the mill from the heating furnace
it is guided into the proper pass by a mill man called a sticker-in.
In addition to guiding the steel into the proper pass, he keeps the
billets from jambing, keeps the different stocks separate and re-
jects any billets too cold to roll. The billet, after being fed into
the first pair of rolls, is automatically passed forward to each
succeeding pair of rolls, and turned ninety degrees between every
other pass by means of suitable twist guides, when it reaches
from one pair of rolls to the next, or by turn-over feed tables
when it comes clear of one pair of rolls before entering the suc-
ceeding pair.
Jan., 1912]
BANTA: WIRE ROD ROLLING
The steel passes from this mill to the intermediate roughing
mill or to the finishing mill, depending upon type of mill used.
If of the Garrett type, it passes to a twelve-inch intermediate
roughing mill, then to the ten-inch train, and is finished in a nine-
inch train and wound up on automatic reels. These twelve-inch,
ten-inch and nine-inch mills usually contain eleven or twelve pair
of rolls.
The 11-8 inch steel billet is passed through cast iron troughs
from the roughing mill directly to the first roll of the twelve-inch
train, is then repeated or guided through a semi-circular trough
to the second pair of rolls, and is then caught by a mill operator
or catcher as it protrudes through the guide on the second pair
Garrett Mill, Old Design.
of rolls, who sticks the rod in the guide of the third pair of rolls,
from which it is passed to the first pair of rolls in the ten-inch
mill, is caught by another catcher who shears off about six inches
of the first end and feeds into the second pair of rolls in the ten-
inch mill, after which it is repeated to third pair and continues in
the same way until it passes the last roll of the nine-inch train and
is automatically coiled on rapidly revolving reels and then carried
by suitable conveyors, narrow guage cars, or trucks, to the wire
mill, or is loaded on standard-guage cars and shipped.
As the rod is reduced in sectional area it increases in length ;
to partially take care of this increase in length each succeeding
train or mill is run at a higher speed and each succeeding roll
Jan., 1912J BANTA: WIRE ROD ROLLING 83
in each train is slightly increased in diameter so as to increase its
surface speed. The twelve-inch mill runs at about 140 to 175
revolutions per minute, the ten-inch mill 300 to 375 revolutions
per minute, and the nine-incli or finishing train 450 to 550 revo-
lutions per minute. The rods will be finished at from 1,400 to
1,550 feet per minute, and will measure over 1,800 feet long
when 200-pound billets are used. It requires considerable skill
on the part of the catcher to catch the rod as it comes from the
mill at 1,400 to 1,500 feet per minute and stick same into the pass
of the succeeding roll, especially as he has to do it very rapidly
to prevent a long loop to be formed on the floor with possible
production of scrap. The diameter of rolls and speeds cannot be
designed to take up all the elongation as produced so loops are
jjermitted to run out on floor each side of each train. Floors made
up of cast iron plates and suitable guides or standings, as they
are called, are laid both sides of mill for each loop to grow in.
These looping floors are sloped away from the mill at about one
in ten so as to permit rod to easily run out on floor. The re-
peaters mentioned, above are so designed that rod will jump
out of groove as soon as the first end enters the succeeding roll.
Such a mill will be finishing from four to seven rods in the last
roll, only one or two in the twelve-inch roll, and only one in the
roug'hing roll. The larger-size mills produce from 150 to 200 tons
per turn or 300 to 400 tons per twenty-four hours.
The section of rods as they pass each roll are alternately
square and oval. The squares are repeated automatically, but the
ovals have to be caught and guided into mill by an operator.
If finishing mills of the continuous type are used, the 1 1-8-
inch steel billet enters the first roll after the first end is sheared
ofif and passes directly to each succeeding roll and to the automatic
reel without the assistance of operators. The billet is also
sheared on the last end as it approaches first roll. The rods are
twisted ninety degrees between every other pair of rolls, same as
mentioned above for the continuous roughing mill, and the rolls
increased in surface speed to take up elongation of the steel, until
the final roll is run at surface speed of about 2,200 feet per min-
ute. It will be noted that this speed is fifty per cent greater than
that of a Garrett mill. Continuous mills are designed to produce
from one to three rods at a time. When finishing these rods, 135
to 190 tons per turn, or 270 to 380 tons per twenty- four hours,
are produced.
The rolls are driven by pinions, cross-shafts, and bevel gears
from main-shaft, the required speed being obtained by using bevel
gears of various ratios, same as on the continuous roughing mill
mentioned above.
84 THE ARMOUR ENGINEER [Vol. 4, No. 1
A third type of finishing mill is sometimes used and is known
as the double Belgian type ; it embodies some good features of
both the Garrett and the continuous mills. It consists essentially
of two pairs of rolls arranged continuously in groups along a
main-shaft, each group being driven at different speeds frorn the
main-shaft so as to take care of the elongation; the first roll and
every alternate roll is turned with oval-shaped passes, and the
balance with V-shaned passes Droducing a square section. The
rod is fed through the twist guide to the second roll in the group,
the same as in the continuous mill- and is then repeated or guided
through a semi-circular trough or repeater the same as on a Gar-
rett mill, and enters the third roll in the mill or the first roll in the
next group, and continues through the mill until finally coiled
on automatic reels at about 2.200 feet per minute, and is then
transferred on suitable conveyors, trucks, or narrow-gauge cars
to the wire mill, or shipped. Such a milh does not require as
many men to operate as a Garrett mill, and no more than
a continuous mill ; further, the scrap is less than on a con-
tinuous mill and possibly a little more than on a Garrett mill.
The Garrett mill is best adapted to roll all sizes, and also copper
rods.
The power required to roll wire rods from four-inch billets
is enormous, beine 6,500 to 7,500 engine horsepower for a mill
producing about 800 tons per twenty- four hours. Each engine
must have capacitv twenty-five to seventy-five ner cent greater
than the average load to take care of extreme fluctuations, as a
rod mill load varies from a friction load to a maximum load at
frequent intervals, and so quickly that it is almost impossible
for bo'ler-house men to keep the boilers from blowing ofif and
still keep steam up. The variations are greatest on the engines
driving the roughing mills.
THE DESIGN AND CONSTRUCTION OF A SEVEN STORY
REINFORCED CONCRETE MERCANTILE BUILDING.
BY E. I. SILVER, C. E.*
The design of a mercantile building demands necessarily the
three following features : First, general arrangement of the
floors ; second, architectural treatment ; and third, structural de-
sign.
The first is governed by the nature of the industries for
whom the building is intended, the amount of shipping provisions
necessary, the divisions of the floors into special rooms as are re-
quired, the division of space for private and general offices, the
amount of toilet room space, etc. The lay-out of the stairways
will be governed both by the occupant and the City building-law
requirements. The City building-laws will determine the width
of stairways, which is dependent on the square foot of floor space
on the various floors.
The architectural features are necessarily governed by the
locality in which the building is to be constructed, the area and
height of the building, the class of industries which are to occupy
the building, and the appropriation of money that can be used to
the best advantage.
The structural design depends upon the fk)or loading require-
ments of the various industries which will occupy the floors, the
arrangement of columns in order to allow the occupants a maxi-
jnum sufficiency in connection with this floor space, and the re-
quirements of the City building-laws.
The building which will now be described both in design and
construction, has just been completed, and is located at the south-
west corner of Adams and Green Streets, Chicago, Illinois.
The dimensions of the building are 125'6'' facing on Adams
Street and 117'0'' facing on Green Street, as shown in the first
floor plan of Fig. 1. The building is of reinforced concrete con-
struction and is of skeleton design so that the brick work need not
progress until the entire concrete work is finished. It is seven
stories in height, making the top of the front coping about 96'
above the sidewalk line. There is no basement, excepting the
boiler room portion, which is shown in the rear and is located
below the raised shipping platform. The height of the first floor
is at the street level.
*Class of li)04. General Superinteiulent, A. S. Alseliuler, Artliiteet, Cliicago.
86
THE ARMOUR ENGINEER
[Vol. 4, No. 1
In di>cussing the design of thi.s building we will divide the
discussion according to the three features as mentioned above.
General Arrangement.
First as to the general plan. The building is to be occupied
chiefly by wholesale tailoring concerns and the columns have been
arranged to suit their recjuirements as is necessary for the lay-out
of cutting tables, sewing machines, sponging equipment ,etc.
The columns as well have been laid out for the most economi-
Figr. 1. First Floor Plan.
cal structural design. The typical panels are l?'/^^" center to
center columns east and west, and 19'lJ/^" center to center col-
umns north and south, and under this lay-out the two-way system
of reinforcing can be very economically used.
The first floor as shown on the accompanying drawing (Fig.
1) is laid out for offices, and shipping and receiving rooms; the
entrance at the corner is used solely for access to the ofifice de-
partment, while the entrance shown on the west is for the various
floors above
Jan, 1912]
SILVER: A CONCRETE BUILDING
87
It is well to note that the various floors above can be sub-
divided into two separate parts, and by installing a partition from
the east line of stairway shaft across the center of the floor to
the east wall of the building, the separate parts can be made ex-
clusive and still have all facilities necessary, such as elevator, stair-
ways, etc. In so doing, however, it will be necessary to enclose
a small space in front of the freight elevator, which is to be made
common to both occupants of the floor, with one door leading to
each part of the floor from this common space.
Figr. 2. Typical Floor Plan.
Another important feature is that the freight elevator can be
used as passenger elevator in the rush hours, if necessary. The
shipping space in the rear of the building is arranged so that the
first floor can ship directly through the two doors as shown and
the various other tenants on the upper floors through the elevator
and shipping floor shown on the north side, making the loading
platform arrangement independently suitable for all floors.
The typical floor as shown in Fig. 2 applies to the sec(«i(I,
third, fourth, fifth and sixth floors.
.S8
THE ARMOUR ENGINEER
[Vol. 4, No.
Installing the stairways, elevators and toilet-rooms in one
location as shown allows the occupant to have his general floor
space entirely together. General toilet provisions have been pro-
vided and are governed by the City and State regulations.
Access to the various floors is provided only through the
doors marked L, which are arranged to be locked, and accord-
ingly each floor can be cut off entirely from the other floors.
The building has been designed for the maximum amount of
JH^^
?7
,
Ck-
^?,
1,
1%
Fig. 3. Top Floor Plan.
light by allowing the windows to take up the entire space between
the columns and extend from the top of the small spandral wall
clear to the ceiling of the floor. Concrete lintels over the window
heads have been turned up, which allows the window to extend
entirely to the ceiling. The building is accordingly well adapted
for manufacturing purposes on account of this maximum amount
of light that can be obtained.
The fire-escapes are conveniently arranged for rapid exit,
complying fully with the Building Laws.
Jan, 1912]
SILVER: A CONCRETE BUILDING
89
The seventh floor as shown in Fig. 3 has been laid out for a
manufacturing floor, and accordingly every foot of space must be
well lighted. There are three large sky-lights in the ceiling as
shown, which introduce the north light only. By using the saw-
tooth sky-light, the intense heat of the sun is kept out, and this
north light, which is most desirable of all, is admitted.
The floors are well ventilated by means of large ventilators
located in the sky-lights.
Fig. 4. Front Elevation, Green and Adams Streets.
Architectural Features.
The architectural design is that of the modern mercantile
building. The two street fronts are a combination of terra-cotta.
brick, and concrete, in conjunction with the special design of win-
dow frames and sash.
Up to the second floor the piers are entirely of brick with a
small amount of terra-cotta trimming, and are tied together by
the heavy band at the second floor, which is topped off with a con-
tinuous terra-cotta sill course.
90
THE ARMOUR ENGINEER
[Vol. 4, No. 1
The entrances, however, are of special design as shown in
Fig. 4, the entrance openings being entirely surrounded with
terra-cotta containing the brick ornaments as shown, and each
entrance is also equipped with two ornamental electroliers at the
transom line.
Above the second story terra-cotta sill, the outside end piers
on each street front are entirely of brick with a small amount of
terra-cotta trimming and are tied together with a heavy brick
band at the seventh floor and roof lines as shown
Fig:. 5. Rear View, Sliowingr Concrete Skeleton Construction.
The intermediate exterior columns from the second tloor line
and base as shown are entirely of concrete exposed to the seventh
floor line where they terminate with a concrete ornamental cap.
These concrete columns were cleaned down and painted with two
coats of cement paint.
The two street fronts are the only exterior parts of the build-
ing which are ornamented, excepting the small pressed-brick
bands around the gravity tank structure.
The remaining two elevations have only common brick with
the concrete skeleton exposed, as shown in Fig. 5.
Jan., 1912] SILVER: A CONCRETE BUILDING 91
The first floor windows on the two street fronts are a com-
bination of store-front and double-hung window design and ac-
cordingly afford a window display in the store front portion if
wanted.
Structural Features.*
The structural details of the building are shown in Figs. 6,
7, 8, and 9. The assumptions, formulae, etc., upon which their
design is based, will now be taken up.
The floor panels, Fig. 6, are designed on a live load basis of
100 pounds per square foot, and the roof slabs on a live load
basis of 25 pounds per square foot. Taking a dead load of 75
pounds per square foot for a six-inch floor panel, and assuming
that the bottom rods on short span take 55% of total load and the
top rods 45%. of total load, we continue to design the panel on
assumptions of formula M=Rhd-. A live load of 100 pounds
per square foot plus a dead load of 75 pounds per square foot
gives a total load of 175 pounds per square foot, which multi-
plied by 55%, gives 97 pounds per square foot taken care of by the
bottom rods, leaving 78 pounds per square foot for the top rods.
Having found the percentage of load in pounds per square
foot taken by the top and bottom layers of rods, we proceed to
find the area of steel necessary to balance the load by first finding
the depth of slab. By using a value of 700 pounds per square
inch for compression in the extreme fibre of the concrete, and
16,000 pounds per square inch for tension in the extreme fibre
of the steel, and with a ratio of steel to concrete of n = 15, and a
percentage of steel of .0073. we derive, by standard formula, a
constant i?= 113.
By using the formula
M=Rbd-
where M is the bending moment in inch-pounds,
i? is a constant, in this case 113,
b is the breadth of beam, 12'',
d is the depth of steel in inches,
and solving for d, we get
-i
M
Rb
*Tlie enyincerinii' desigiiiiis' of this bnililiug- was done by G. 1'. Claysoii. C. E.,
of A. S. Alschulei-'s office.
92
THE ARMOUR ENGINEER [Vol. 4, No. 1
'igr. (i. Tyidcal Hoiiiforced Concrete Tloor Panel.
94 THE ARMOUR ENGINEER [Vol. 4, No. 1
Substitutintr values, we have
|97X17.5X17.5X12
113X12X12
= 4.7 inches to center hne of steel.
For total numlier of bars in span,
area of bar
Substituting the value of p = .0073, percentage of steel, and using
Yj" square bars, we get for a span of 17'/^'" (practically 18')
4.7X12X.0073X18X7
A^ = = 22 bars.
.25X9
The constant. 7/9, is taken from Turneaure & Maurer. rnd
explanation can be found in that text-book, as the discussion is
too long for this article.
For top layer of rods d = 4.6" and X = 22-y2" rods, calcu-
lation being similar to that above.
In determining the total thickness of panel we add 4.6".
which is distance from top of slab to center line of top layer of
rods. .7" between centers of bars. .35" from center of lower
bar to edge, and .5" from lower edge of bar. to under side of slab,
which gives us. for practical purposes, a 6^^" panel.
One-half of the rods are placed in the center third of the
panel and one-quarter of rods in each remaining third, spacing
the center third 6" on centers and remaining rods ecjuidistant to
beams. Since the panel has been figured as continuous over two
supports, using a bending moment of WL/12. the same number
of top rods is put over the support a distance of two-fifths of
the span to develop the full strength of steel at this point, as
the bending moment at support is the same as at the center
line of panel.
For end panels proceed as above outlined except in place of
bending moment JVL/12 use WL/IO.
The beams or girders, Fig. 7, were figured as T-beams. using
a value of 700 pounds per square inch in compression for the
concrete and a value of 16.000 pounds per square inch in tension
for the steel. Using the formula M ^JVL/12 for the bending
moment at center line of beam, or M = WL/6 for the total
Jan., 1912] SILVER: A CONCRETE BUILDING 95
moment at the center line of beam and support, the bending
moment M is found for use in the formula for T-beams.
M
where .-^s is the area of the steel in square inches,
/s is the fibre stress in steel,
d is the depth of steel,
.86 is the ratio of arm of resisting couple to d.
Assuming the beams 10" by 20" we find the loading, which
for a 19'1><2" span is, under a load of 97 pounds per square foot.
19'lK'"Xl7'/'y2"X97 = 34,000 pounds,
to which add 4,000 pounds for the weight of the beam, whicli
gives a total load of IV = 38,000 pounds. Therefore
38,000X19X12 .,-^.. . , ,
M=^ : = /2o,000 mch-pounds.
12
and
725,000
16,000Xl8X-86
3.1 square inches.
Use two ly." by 2j4" Kahn bars, and two 3/<" bars, these giving a
total area of steel of 3.3 square inches.
Since the beams are figured as continuous, put the same
amount of steel over support, using two 1^" by 2}^" Kahn bars 7'
9" long, and bending two '/>" square twisted bars over supports
for the same distance. For short span beam proceed in the same
manner, figuring load as
19'li4"Xl7'7>V'x78 = 26,000 pounds,
to which the weight of beam (4,000 pounds) is added, giving a
total load of 30,000 pounds. End beams are to be figured ac-
cording to bending moment of IVL/IO and the load above given.
The columns. Figs. 8 and 9. were figured in accordance with
the Chicago building ordinances, scaling down the live loads per
square foot 5% for each floor, starting with 85% on the seventh
floor, and adding the entire dead load each time. The lower
columns were figured as hooped columns, in order to keep them
as small as possible, and 1-1^.-3 concrete mixture was specified
for first, second, and third floors, while for the remaining floors
a 1-2-4 mixture was specified.
v^^
^\',^
^M-^^
^i4
>
v;
v5fe
7~^
•lu^
^r^^
^i>'^
,'>'^'^^>^«
^.5|'
^"^i
^^.^
Arrse/o^e Coi-
^1
Jan., I'Jlil SILVER: A CONCRETK BUILDINCJ
97
lig. 1). Columii Sections, Footings, and Pent-house ('oustriution.
THE ARMOUR ENGINEER | Vol. 4, No. 1
The footings, Fig. 9. were figured in accordance with stand-
ard methods, using a high elastic-Hmit steel, 1-3-5 concrete mix-
ture, and a soil bearing pressure of 3,500 pounds per square
foot. The allowable bearing pressure of top plate under column
was figured at 350 pounds per square inch, and a shear value on
a diagonal plane of 40 pounds per square inch was used.
Construction.
The amount of excavation was not very great for the reason
that there was no basement to the building. The only excavation
required was for the boiler footings, the boiler room space and the
])ier and wall footings, which went down to hard blue clay at a
depth of 7'6".
The retaining walls as shown around the boiler room were
constructed entirely of concrete mixed to the proportion of one
part of cement, three parts of sand, and five parts of gravel.
The pier footings throughout the work w^ere constructed of the
same mixture. The steel used in the footings w^as of the high
elastic limit twisted steel giving an elastic limit of 55,000 pounds,
and was tested by a reliable testing laboratory before acceptance.
The soil upon which these footings rest was of the very best
blue clay entirely clear of any loam or water. Borings and tests
were made before the building was erected and this same kind of
clay was foiTnd continuing on to hard sand.
The reinforced concrete work of the building proper started
at the top of the column footings, and continued on throughout
the remainder of the building. The building is of skeleton con-
struction, so that the columns, beams, slabs and girders were en-
tirely constructed before any of the brick work was started.
On account of the great amount of concrete in the building.
the walk space and part of the roadway on the Green Street side
were utilized for depositing material. Oh this side the mixing
was done, the concrete being dumped into a large bucket, and
hoisted up in a tower to the respective floor on which the pouring
was being done, as shown in Fig. 10.
Run-ways were built upon the floor throughout to allow for
the traveling of carts, which transferred the concrete from the
bucket to the portions of the floor being poured. Before the
pouring was started the necessary sleeves for heating, plumbing,
and wiring and machinery hangers were installed. The rods were
then wired into place and set on concrete blocks to keep them the
required distance from the deck. Sprinkler hangers were in-
stalled later with an electric drill. The wiring was all run in
conduit which was placed on the wood decking before concreting,
and then concreted in with the rest of the work .
Jan., 1912J SILVER: A CONCRETE BUILDIXG 99
The concrete used in the reinforced concrete work was mixed
one part of cement, two parts of sand and four parts of gravel.
The gravel was preferred to crushed stone on account of the bet-
ter fireproof quality.
The floors throughout were finished concrete Avith cement
top ; accordingly the finish had to be installed at the same time that
the remaining part of the floor was poured.
The first three inches of floor were poured very wet in order
that the concrete would fill the entire space underneath and about
rig. 10. Building Under Construction.
the rods, insuring in this way smooth ceilings, columns, and gird-
ers throughout, requiring little pointing afterwards.
The top three inches of the floor were poured very much
drier and tamped so the water just came to the surface. In this
way the finish could be put on nearly as soon as the required
depth of concrete had been poured.
The top rods which occur over the beams were not set in
place until after the rough concrete had been poured, when they
were then tamped in so that the top of the rods would be just
underneath the finish. The steel used throughout the floors was
the same kind of twisted steel as described for the footings.
ini) TIIL: AirXlOL'R EXGIXEER IVol. 4, No. 1
The starting" of the reinforced concrete pouring was on June
22, 1911, and the pouring of the roof, excluding the gravity tank
and sky-Hght work, was finished on September 19, 1911 ; thus a
floor was poured every ten days, which was very good progress
considering the fact that work was done through only one shift.
When the roof was ready for pouring the brick work on the
outside started at the grade line and the entire brick work wa.-^
completed, excepting the tank house, on November 15, 1911, with
fifteen days delay on account of rain and cold weather.
Auxiliary Equipment.
The heating plant installed is a low-pressure system with two-
pi])e vacuum installation. The boilers are of the horizontal tubu-
lar type with smoke furnace in accordance with the rules and reg-
ulations of the vSmoke Department of the City of Chicago. The
main is carried around on the ceiling of the top floor and the re-
turn is located in a concrete trench beneath the ground floor,
pitching back to the boiler with the proper pitch as necessary.
The building throughout is equipped with standard radiators and
the heating system is designed to heat it to a temperature of sev-
enty degrees when the temperature outside is ten degrees below
zero.
There is also installed a sprinkler system in accordance with
the rules of the Protection ?*Iutual Fire Insurance Company and
the Chicago Board of Underwriters. It consists of the under-
ground work, risers, pipe-lines throughout the various floors,
s])rinklers, pressure tank, gravity tank, etc. The approximate
total number of sprinkler heads will be 1,250 arranged in panels
and installed on the "wet" system. The gravity tank has a capacity
of 20.000 gallons and the pressure tank a capacity of 4,500 gallons.
The air compressor and filling pump are located in the boiler
room and are electrically driven. Bells are installed on the various
floors and annunciator in the boiler room, so that at any time that
the heads are discharged, it is reported back into the boiler room.
On account of its fireproof construction and sprinkler equipment
this building has the lowest rate of insurance that the companies
allow. The building is now in its final stages of completion and
will be occupied throughout by January 1, 1912. It was designed
and constructed under the direction of Alfred S. Alschuler, '99.
Architect, 1900 Stcger Building, Chicago.
THE ROLE OF THE AERODYNAMICAL LABORATORY IN
THE DEVELOPMENT OF AERONAUTICS.
BY SYDNEY V. JAMES, M. E.*
Now that the aeroplane is becoming recognized as a prac-
tical machine, the interest of the scientific and technical world is
being turned toward its development more and more, with the
result that laboratory methods are being utilized and systematic
results obtained. Such laboratory work is most useful from the
point of view of the designer, and when undertaken systematically
m conjunction with the results of actual practice, much can be
accomplished with small expenditure of money and time, not to
mention the reduction of the toll of human lives which has been
])aid in the past.
The history of naval architecture was marked with much un-
necessary expense and loss of life until the designer had his work
made more definite for him by carefully conducted experiments
on the powering, strength, and rolling of ships. This work is
now being carried on to a great extent in laboratories entirely de-
voted to such work. One of the most useful departments of these
laboratories is the towing-basin, in which scale-models of ships
arc tested and their resistances at different speeds carefully meas-
ured. By comparisoii with tests on full-sized ships of various
kinds, the necessary coefficients, and ratios of the model-results
as compared with tests of the full-sized ship, become definitely
known. This method can be, and has been, applied to aeroplanes
to a certain extent, but of course the importance of such work has
not yet made itself felt among nations to such an extent as has
the naval work.
In following the above analogy into one of its details it may
be said that ship-model testing for resistance is more complicated
than the corresponding work for aeroplanes. This is true, be-
cause, first of all, the ship-model must be towed at strictly scale-
speeds on account of the fact that the relations between resistance
and speed in such a case are very uncertain. By "scale-speed" is
meant that according to the law of comparison, the speed of the
model is to the speed of the ship as the square root of the
length of the model is to the square root of the length of the
ship. Another reason for the complication of ship-model testing
H'hiss of l'J07. Aeronautic Kiisiiiepi- f„r n.-iiold F. McCoriuick, Harvester
r.uildins, Cliicnfio, Illinois
102 THE ARMOUR ENCxINEER [Vol. 4, No. 1
lies in the fact that the effects of wave-making: and skin frictional
resistances must be separated from each other in the computation
of results. This complication does not affect aeroplane models,
because between the limits of practical velocities of flidit the re-
sistance varies practically as the square of the velocity, hence the
results obtained under one set of speed conditions may easily be
transferred to any other set of conditions, and the relations be-
tween small and large scale surfaces and objects as to resistance
and other aerodynamical effects are being established more and
more definitely ; so that results are transferable from one scale
to another.
Let us consider a little more definitely just the kind of work
being accomplished and the principal methods in use at the pres-
ent time. Tests are being made to determine the resistance to
horizontal motion, the vertical or weight-opposing component, and
the location of the so-called center of pressure for all types of
lifting surfaces so that the results may be used for designing an
aeroplane or other aerial vehicle. The resistance to forward mo-
tion, and hence the determination of the form of least resistance,
of all framework, fuel tanks, radiators, and engines associated
with aeroplanes or dirigible balloons, is to be carefully determined
and made available for engineers. Propellers come in for their
share of the work. Alost interesting results are now being ob-
tained by comparison tests of propellers built to the same design,
but of different scale, for the purpose of establishing a basis upon
which to predict in the future the behavior of a given propeller
very accurately from tests of its small scale model operated under
known conditions.
There are several methods open to use for obtaining these
results, but the most important are: (1) The testing of small
scale models suspended by sensitive balances in a current of air
generated in a so-called wind-tunnel; (2) The mounting of larger
scale models on a car which may be driven along a track, prefer-
ably straight, in the open air; (3) The mounting of a small or
large scale model on the end of a whirling arm. In all such equip-
ment very sensitive means are available for determining the veloci-
ties, pressure, power applied, etc.
Let us consider as a typical example of an installation in
which the first method is utilized, the aerodynamical laboratory
of M. Eiffel in Paris. Fig. 1 shows a cross-section through the
building, the essential parts of which are the room A, from
which air is drawn through the bell-shaped inlet B, through the
experimental chamber C. and into the motor-driven exhauster
D. The air then rcLiu-ns to the room .1 whence it started, thus
Jan., 1912] JAMES: AERODYNAMICAL LABORATORIES
103
completing the circuit. The current of air passing through the
chamber C is rectified and made as uniform in velocity through-
out its cross-section as possible. This is accomplished by means
of the rectangular cells in the intake to the chamber at B, and
the netting in the intake to the exhauster shown at F.
The model G to be tested is suspended from the balance H
and an observer above in the upper room /, makes the necessary
measurements of the forces acting while a current of air passes
through the experimental chamber below him. The velocity of
the air is measured by means of Pitot tubes. It is interesting to
note in this connection that several methods of measuring the
velocity of the air were tried out. These were: (1) by Pitot
tube; (2) from the difference in pressure between the air out-
side and inside the experimental chamber (which under some
^^'— j 1
1
....
-1 -
1 —
^ A ^-
\
i
7
/ '.r^-— i
1
ii
^
j.
4
'IQ
ft
y
„
A
■^
w
1 _B_
1
A
\
1 :
1
i
Fig. 1. Eiffel's I.aboratory.
conditions reached as much as twenty millimeters of water) ;
(3) by different forms of anemometers. The calculation in the
second method was by the means of the formula J'-^2gh~d
where h is the difference in pressure as measured by the water
barometer, and d is the specific gravity of the air. The second
and third methods checked to within one or one and a half per
cent of the Pitot-tube readings.
Fig. 2 is a reproduction from a photo showing a small scale-
model of a Nieuport aeroplane being tested in the current of air.
Fig. 3 gives a more general view in the experimental chamber,
showing on the right the air intake cells, B, previously men-
tioned, as well as the observer above. On the left is seen the
exhauster intake.
A brief outline of the work accomplished in this laboratory
will now be taken up with a view to showing the thoroughness
as well as the comprehensiveness of the methods in use.
Till' .\UM()L:R engineer |Vo1. 4. X(.. 1
Jan. 1912] JAMES: AERODYNAMICAL LARORATORTKS li
(1) A series of tests to determine the effect of size of
surface upon the pressure per unit of area. This was carried
out with the air moving normally to the model, and the resist-
ance as represented by the equation. R=KSJ'^-, determined. A.'
is the coefficient. .S' the area, f the velocity. When R is ex-
pressed in kilograms, 5^ is sc|uare meters, and V in meters per
second. K was found to range from 0.065 for an area of 0.01 of
a square meter to 0.08 for an area of 1.0 square meter, and
showed a tendency to become practically constant for areas above
Illustration from Scientific American.
Figr. 3. General View in Experimental Chamber.
the latter amount. These tests were made using square plane
surfaces.
(2 ) A series of tests, also with plane, normal surfaces, but
to determine the eff'ect of the ratio of length to width, or what
is called "aspect ratio." Surfaces of 225 square centimeters
were used and the aspect ratio varied from 1 to 50. The values
of K showed an approximately uniform increase from 0.065 to
0.096, the last figure being 509; higher than the first.
(3) Planes and curved surfaces of various aspect ratios
106
THE ARMOITR ENGINEER [Vol. 4, No. 1
were tried, set at different angles to the current, and the values
of K determined as before. These values were elaborately plot-
ted, showing their ratio to the K for normal incidents of air
current. The curved surfaces showed marked advantages by
giving much greater ratios at the small angles of incidence.
(4) Another series of tests was made to determine the po-
sition of the center of pressure on the surfaces at different angles
to the current. These data are extremely useful in practical work
as they enable the stability problem to be more accurately studied.
Separate series were undertaken to show the effects on the cen-
ter of pressure of various amounts of curvature for surfaces
such as are used in large aeroplanes.
(5) A most interesting set of measurements was taken of
the distribution of the pressures both on lower and upper sides
of an ordinary surface at various inclinations to the current. It
was shown that the "partial vacuum" acting on the upper surface,
at ordinary flight angles of inclination, exerted nearly 70% of
the total upward lift.
(6) The ratios of upward or weight-carrying component,
called the 'Mift," to the resistance in line of motion component,
called the "drift." for models of all standard aeroplane surfaces,
were determined under all conditions. Some of these tests
showed the effect of interference between surfaces in the case
of an ordinary biplane.
(7) Complete scale-models of aeroplanes (to 1-10 scale)
were suspended in the current and some very interesting experi-
ments leading to power calculations were performed. (See Fig.
2.) The results were in wonderful agreement with full-sized
practice when all necessary allowances were made in referring
them to full-sized scale.
(8) Propellers formed the subject of another series of
tests in the wind-current, and gave remarkable results, bidding
fair to establish a good working basis for future testing of pro-
posed propelling devices. A particular series showed the com-
parison between a model of a "Normale" propeller and the full-
sized one, and the results when correctly interpreted are in very
close agreement.
Other installations which carry on work in accordance with
the first of the above-mentioned methods are the aerodynamical
laboratory at Koutchino in Russia, the aerodynamical laboratory
at Goettingen in Germany, and the National Physical Laboratory
at London, England.
As a typical example of an installation utilizing the
second method of attack, we shall choose the Acrotcchnical
Jan., 1912] JAMES: AERODYNAMICAL LABORATORIES lO^^
Institute of the University of Paris at St. Cyr in France.
This institution was founded by the generosity of M. Henry
Deutsch. Its object is entirely scientific, and is to study all
theoretical and practical problems of aviation and aerostation.
Its equipment is very complete and covers an area of some eight-
een acres. The main building contains several wind-tunnels de-
signed for various purposes, aerodynamical balances, thrust-
measuring devices for propellers, chemical and physical labora-
tories, and photographical department. In the grounds there is
an experimental track about 4,600 feet long, perfectly straight
and level except for 250 feet at one end, which has a slope of 1
in 100 to assist in starting the rolling platforms, and a rise of 1
in 20 at the other end to aid in stopping and returning them. The
rails are 36 feet in length and welded two together so as to give
72 feet without a joint. The current supplied to the electrically
driven cars running on this track is supplied through two "live"
rails, one on each side of the track, mounted on posts about two
feet high. The current returns through the track rails.
Four rolling platforms or cars are provided, each being de-
signed for carrying out some particular branch of research. One
is fitted to measure and record the horizontal and vertical com-
ponents of the air pressure on all kinds of surfaces, as well as
determine the center of pressure, for various angles of incidence.
Another is designed to measure resistances of various objects in
motion through the air. The other two are for propeller testing,
one being fitted to carry large-sized propellers such as used on
dirigible balloons, and the other for ordinary-sized aeroplane
propellers. Both these cars are,, fitted with instruments for
measuring thrust, power supplied and used, and speed of rota-
tion.
Fig. 4 is reproduced from a photograph showing the first
car above described. A plane is shown mounted for test. (The
main buildings are to be seen to the right in the background.)
This car weighs, including the motor of 500 pounds, 2,220
pounds. The steel frame is twenty feet long and six feet seven
inches wide. This frame is carried on two axles eleven feet nine
inches apart and projects six feet in front of the front axle and
two feet three inches in rear of the rear axle. This difference
is necessary to put enough additional weight on the front axle,
over which the lifting surface is situated, so that in testing large
surfaces, the front of the car will not be lifted. Four ball-bear-
ing mounted wheels, three feet three inches in diameter, carry
the car. The speed attainable is about 110 feet per second or 7t^
miles per hour.
lOS
THI'. ARMCWR ENGINEER
[Vol. 4, Nn. 1
The instruments fitted comprise: (1) a registering chrono-
graph recording axle revohitions: (2) an independent speed
registering device; (3) pressure indicators for measuring lift
and drift of the test surface; (4) wattmeter for determining the
power used throughout the run.
A vast amount of valuable work has been done with this
equipment. As surfaces up to eighteen or twenty feet in span
by about six or eight feet fore and aft dimension can be tested,
the exactness of the results enables us to calculate very closely
Illustration from Scieutitie American-
Fig. 4. Open Air Test-car at St. Cyr.
the relations between very small models, such as used by Eififel
and others, and those of nearly full size. The result of these
opportunities for comparison and correlation is to rapidly fill
the gap in this branch of the work, and place engineers one step
further toward a rational basis of design.
Work along this same line of straight track, open air test-
ing has been carried on by a number of other experimenters.
Jan., 19121 JAAJES: AERODYNAMICAL LAHoRATUKIliS 109
W. R. Turnbull, AT. E., of Canada, has conducted a series of
experimentst with screw propeUers ranging in diameter from 1.5
to 3.5 feet, mounted on a car with the proper instruments for
making the necessary observations. His results are of consid-
erable value. He has shown, for example, that the proper pro-
portion of the pitch of the aerial screw propeller to the diameter
should be in the neighborhood of 1.35 in order to get maximum
efficiency with minimum slip. This indicates that the Wright
brothers, having selected a ratio of 1.2 and then having geared
down the propeller to its best speed, showed an appreciation of
the actual conditions of the problem. Their results are remark-
able from the point of view of efficienc)-, and indicate that fur-
ther progress along this line may be made by careful laboratory
work checked against their field tests.
The third method, the whirling table, is perhaps the oldest
form of experiment used to study air resistance, and is still used
on account of its simplicity and convenience. It consists in gen-
eral of the object to be tested, mounted upon the end of an arm
of considerable length (compared with the size of the object)
and swir.ging the arm in a horizontal plane about a vertical axis
through one end. Chronologically the method dates back to
Robins, 17-16, and Hutton, 1787. both of whom performed some
very interesting experiments. Dines in England has used this
method and Prof. Langley of the Smithsonian Institute, Wash-
ington, established a large number of the fundamental laws of
aerodynamics by means of a whirling table.
A typical example of this form of table is shown in Fig. 5,
which is a drawing of the whirling table in the National Physical
Laboratory, London.* The arm has a radius of 30 feet, and is
constructed of steel tubing tapering from 1^ inches at the axis
to 1 inch at the end. and connected together by cross struts. The
whole is strongly braced with steel wire. The tubes are IZV^
inches apart and suspended from a cantilever truss built on the
center post. This center post is supported at its upper end by a
bearing fastened to one of the roof trusses. A 14-horsepower
electric motor drives the table by means of worm gearing which
reduces the speed in the ratio of 28 to 1. In order to avoid dis-
astrous results to the motor and gearing due to inertia of the
arm, the post is cut above the worm wheel and the parts con-
nected together by a ratchet gear. This allows the arm to con-
tinue moving when the motor is stopped. The rotating speed of
■j-"Aeronautical Journal." London, January, 1911, page 20.
*See "Report of the Advi.sorv Committee on Aeronautics," 1000-1010, London,
1010, papre 15,
no
THE ARAIUUR KX(iL\Kl£R LVol. 4, No. 1
Jan., 1912] JAMES: AERODYNAMICAL LAIU UMTOUIKS
tlie arm may be varied from 5 to 30 revolutions per minute,
which means that the Hnear velocity at the tip varies from 10 to
60 miles per hour.
The arm has been specially fitted for testing small propellers.
The figure shows a one-half horsepower motor on the arm at a
distance of eight feet from the end, driving the propeller shaft
shown at the end of the tubing. The mounting of this shaft is
very ingenious and embodies means for measuring the thrust,
as well as the torque and the number of revolutions per minute.
The shaft is mounted in ball-bearings and is allowed a small
amount of longitudinal movement. This movement is con-
trolled by a tension spring brought into operation through a
linkage. The torque is measured by driving the shaft through a
spring-controlled coupling. A pencil draws a line on a drum and
shows the extent of the motion both for torque and thrust. A
small direct-connected generator connected with a voltmeter on the
stationary observing platform through slip rings on the center post,
gives the reading for the number of revolutions of the propeller.'
The linear speed through the air and the revolution speed of the
propeller may be controlled independently of each other. Any
other apparatus for testing air pressures or other forces may be
mounted on the arm.
Some disadvantages of this method are: (1) that the whirl-
ing table if placed in the open air will be affected by the atmos-
pheric conditions; (2) that if placed in an enclosed space it will
set the air in that space in motion, causing an error in the deter-
mination of the velocity of the model through the air; (3) the
model will be interfered with by eddies formed by the revolving
arm. On account of these objections the British Government
Committee believes that the method is not susceptible to such
refinement as the wind-tunnel method.
The principal advantage of the method is. of course, its
convenience, and if used in a closed room the proper calibration
can be made and the allowances for circulation of the air in the
room as well as other sources of error can be applied. Such cor-
rections are available for the above described apparatus, and some
very valuable results have been obtained in the way of compara-
tive tests of small- and large-scale propellers. A very interest-
ing example of the application of the law of comparison which
was made there is the following. A large propeller made by
Messrs. Vickers, Ltd., and tested on their large whirling table at
Barrow, had a diameter of 15 feet and its test showed the fol-
lowing results : —
112 'I'lIK AK.MOL'k JiNGlNEER |Vul. 4, No. 1
Speed of translation I '^=36 miles per hour.
Propeller rcvolution> V^=450 per minute.
Thrust 7 =864 lbs..
Efficiency E^^M ' ', .
Horsepower aljsorbed . . H F=130.
X;nv according to the law above mentioned, tlie ratio of the
speeds of trai'.-lation of the model to that of the large propeller
sliould be in projiortion to tlie s(|uare root of the ratio of the linear
dimensions. Following this basis the data were calculated for a
model two feet in diameter. .\ model of such size was made and
tested and the table below shows tlie figures predicted according
to the law of compari.son and the actual test results obtained at the
National I'hvsical Laboratory on the whirling table: —
Pounc
Is 'I'hrust
Horsepower
Efficiency
'alues
deduced
from law
2.05
0.1123
64.09^,
,'alues
obtained
by test
l.<)7
0.1115
61.9%
This ver_\- clearl\- indicates the yalue of carrying out more of
such work and proving that the model basis is extremely valuable.
The field of work that has been covered l)y all kinds of aero-
dynamical laboratories is large, but the possible range is so great
that there is need for the establishment of a fully-equipped test-
ing i)lar.t in tliis countr_\' to be devoted t(~) this work. It would
simply mean that tiie science and art of ax'iation would be ))laced
upon a reliable working liasis without unnecessary waste of time
and money, and much work could be carried out without risk tn
human life.
LOW TENSION FEEDER SYSTEMS FOR STREET
RAILWAYS.t
BY RALPH H. RICE, E. E.*
An electrical distribution system in general is made up of a
network of conductors which convey the electrical energy from
its points of generation to the locations at which it is to be util-
ized. The distributing lines for a street railway system consti-
tute a special and relatively simple case of such a network. Its
component parts are :
1. The contact conductor, or trolley wire, from which the
car receives its current.
2. The conductors, or feeders, from the power house or
substation switchboard which deliver current to the trolley wire.
3. The returns, or conductors, which complete the electrical
circuit from the car to the power station.
It is the purpose of this paper to explain the methods used
in calculating the feeders and to show the general results ob-
tained on the street railway systems of Chicago, which come
under the jurisdiction of the Board of Supervising Engineers.
There are four companies with 687 miles of single track oper-
ated, and a total of fifteen power stations, aggregating 95,150 kw.
rated capacity. The calculations for feeders were made upon
a basis of 2,264 double truck cars, or their equivalent, which
were required to operate the proposed schedule.
On the lines of these companies but two steam power houses
are in operation. In all other cases power is purchased and is
transmitted at 9,000 volts, three phase, to the various substations
and there transformed to direct current. All of these substations
have been very recently built and are of the most approved con-
struction throughout. At present the direct current bus voltage
is run somewhat lower than 600 volts, but it is intended to raise
it to this value, after low voltage motors have been eliminated
and other necessary changes have been made.
The trolley wire on all streets is sectionalized by inserting
♦Class of 1897. Ass't Engineer, Board of Supervising Engineers, Cliicago Trac-
tion Company.
f A paper published in the JOURNAL of the WESTERN SOCIETY OF ENGIN-
EERS, June, 1910; awarded the Octave Chanute Medal as the best paper on
electrical engineering presented during 1910 before the Westerq Society of
Engineers,
114 THE ARMOUR ENGINEER [Vol. 4, No. 1
section insulators at various points, and each of these trolley sec-
tions has a feeder consisting of one or more cables extending
from the section to the power station most convenient to it. The
choice of location for these section insulators is largely dependent
upon operating conditions, and also upon the location of a section
with reference to the power station.
To determine the actual power requirements for the type of
double truck car now used in Chicago, a series of fifteen tests
were made on a car which was operated in regular service for
three days. This car was equipped with ammeters, voltmeters,
a recording watt-hour meter, and a speed recorder. A careful
log was kept of its operation under all the varying conditions of
service on different lines and at ditTerent periods of the day.
In addition to the above individual car tests, several section
tests were undertaken. These consisted in determining the num-
ber of cars on a chosen trolley section by observing the times of
entrance and exit of cars from the section and during the period
of the test, noting the ammeter and voltmeter readings at the
station switchboard. The chosen sections were of several miles
in length and the traffic over them was quite heavy. All tests
were run between 12 m. and 8 p. m. From the results of these
tests the power requirements for a car under various conditions
were determined and used in the feeder calculations.
Some study was also given to the safe carrying capacity of
cables, and as a large portion of the territory covered by the
feeder system is designated as underground territory by the or-
dinances under which the railway companies are operating, it
was necessary to consider lead sheathed, as well as weather-proof
overhead cables. Owing to the presence of various tunnels, low
subways under railroad tracks, and some low level duct lines, it
was considered desirable to use rubber insulated cables in these
places and paper insulated cables elsewhere.
A question of considerable importance is the interconnec-
tion of stations so that the temporary shutdown of one will not
throw out of service all the sections fed from that station. One
method of accomplishing this, would be to install into every
station sufficient copper leading from the adjacent stations to
enable the full-current capacity of the disabled station to be fed
into it from the others. The presence of such trunk tie lines
would be highly desirable during an emergency, but would be
of no service at other times, except as equalizers between the
stations. The investment necessary for such ties would be quite
large, and essentially the same results may be secured by select-
ing a certain number of the more important trolley sections and
Jan., 1912] RICE: LOW TENSION EEEDER SYSTEMS
115
feeding them from two stations. The feeders to these sections
are so proportioned and calculated, that on the whole system in
case of the shut-down of one or two stations, a certain propor-
tion of the cars can be carried on the remaining stations by in-
terconnecting through these tie lines.
On any individual trolley section, in case of the shut-down
of one station, or of accident to one feeder, the cars on the sec-
tion could still be operated from the other station by means of
the second feeder.
NOT TO SCAUe
Station Capacity in Amperes.
The illustration. Fig. 1, shows the station capacity of one
system in amperes, for operation at 600 volts. Local conditions
and the situation of the stations determine the tie capacity be-
tween any two of them. In the system here shown the stations
are very completely tied together and each has a tie capacity into
it, which is a large percentage of its output.
In any street railway system the load will fluctuate widely,
and it is a question of importance to determine upon what par-
ticular basis to design the feeder system. The load factor, that
116
THE ARMOUR ENGINEER
LVol. 4, No. 1
is, the ratio of the average load during any period of time to
the maximum within the period, is practically 507o for a twenty-
four hour period on the systems under consideration. The curves
of Fig. 2 are typical for a twelve-hour period. The other twelve
hours give practically duplicate curves.
A — December '07 Maximum.
B — January '08 Maximum.
C — January '08 Average Maximun
D — 3 3Ionth8 Average Maximum.
E — 2-Hour Average.
r— 4-Hour Average.
G — 6-Hour Average.
H — 13-Hour Ayerage.
The average maximum curves C and D represent the aver-
age of maximum daily values for the selected months. The
average curves for the various number of hours, such as B and
F, represent the steady load which would be equivalent in kilo-
Jan., 1912] RICE: LOW TENSION FEEDER SYSTEMS 117
watt hours to the actual load as it exists on the station for the
twelve-hour period shown.
If the feeder system is designed to carry the peak loads, such
as A and B, without an overload on the cables, then during a
large part of the day the current in the cables will be far below
their safe carrying capacity and considerable copper will be idle.
On the other hand, if the feeder system should be designed on
a basis of say a six-hour average, as shown by curve G, then
the cables would be subjected to large overloads for a consid-
erable period of time, heating would occur, the cables would more
rapidly deteriorate, and a shorter life would result.
It was decided to use the two-hour average as a basis for
all calculations, in this case represented by curve B. This means
that the feeder systems are so designed that they will carry
the entire load without being overloaded, except during a two-
hour morning and evening peak. The ordinary percentage of
excess load is well within the overload capacity of the cable —
that is, the current it will carry for this short period without
undue heating.
From the preliminary studies and tests made by the Board
of Supervising Engineers, as previously outlined, the basis for
all feeder calculations was formulated and established by reso-
lution. This basis may be itemized as follows :
1st. — The direct-current bus bar at power houses or sub-sta-
tions wnll be operated at approximately 600 volts.
2d. — An allowance of 40 kw. in power house or sub-station
capacity for each standard double-truck car of the type
approved by the Board of Supervising Engineers,
weighing approximately 26 tons, light, or its equiva-
lent, will be provided at each direct-current bus bar.
3d. — In calculating the copper for current-carrying capacity
an allowance of 75 amperes for each standard double-
truck car, as described above, or its equivalent, shall
be allowed.
4th. — An average drop of 50 volts will be allowed between
the D. C. bus bars and the center of gravity of the
trolley section, due provision being made for suitable
tie lines to take care of emergency cases.
5th. — The carrying capacity in amperes of insulated lead-
covered underground cables and of overhead weather-
])roof ca])les shall be calculated upon tlie following
basis :
lis THE ARMOITR ENGINEER [Vol. 4, No. 1
Lead Covered Triple Braided
Rubber Paper Weatherproof
1.000,000 CM. Cable 800 1000 1250
500.000 " " 500 600 625
350.000 •• " ^7S 425 450
4/0 " ... 325
The general method adopted in feeder calculations may be
illustrated by the successive steps used in one particular system,
the detail of calculation being shown later.
1. From the proposed schedules, the number of cars which
will be operated on each route during the rush hours is deter-
mined. The cars are distributed and plotted upon a skeleton map
of the system which is called a Spot Map or car distribution
map; Fig. 3. The afternoon maximum period is usually the
heaviest service period, so that tlie car distribution for two hours
of what is styled the P. M. Rush is used on this map.
2. The trolley sections, as previously determined, are then
drawn, and the number of cars on each are multiplied by 75,
which gives the total average maximum load Tor each individual
trolley section in amperes. This amount is placed in a small
circle at the center of gravity of the trolley section, and the map
is known as the trolley section or load distribution map, as shown
in Fig. 4.
3. A study is then made of the proper location of power sta-
tions. The best probable locations are selected, and a calcula-
tion of station load centers is made by finding the combined
center of gravity of the loads about a given station, as indicated
in Fig. 5. This center of gravity of the loads is determined by
the well-known process in mechanics, in this case the number of
amperes on each section taking the place of the number of
pounds weight. In the case of tie sections, the load is divided
between two stations in amounts which may readily be cal-
culated, as shown later. The section is thus virtually ctTvided into
two sub-sections of such lengths that the load on one sub-sec-
tion is carried by one station and the load on the other is car-
ried by the second station. The dotted circles represent the cen-
ters of load of these sub-sections, and the numbers within them
the portion of the total on the sub-section.
If a given system is to be fed by a single power house, the
system load center is also determined, whicli will show the most
economical location, so far as distribution copper is concerned,
for the generating station. If the locations chosen are not the
most economical for distribution copper, studies are made of com-
Jan., 1912] RICE: LOW TENSION FEEDER SYSTEMS 119
parative cost for other locations where the company may have
property or where real estate for sub-station purposes may be
obtained to advantage.
4. After the station locations are definitely settled, and the
sections which are to be fed from each station are decided upon,
a Spider Diagram is added to the trolley section map. which now
becomes a drawing of record and shows at a glance what sec-
tions are fed from any given station, and what average maximum
load is to be expected upon that section. This is illustrated in
Fig. 4.
5. A study is then made of the feeder routes, and having de-
termined them, a Feeder Diagram is prepared which shows the
route and number of each cable from the power station to the
section load center, as shown in Fig. 6.
6. If the feeders are to be placed underground, it is neces-
sary to lay out conduit lines. A diagram, Fig. 7, is used for
this, the number of cables over a given section being repre-
sented arbitrarily by the numerator of a fraction, and the num-
ber of ducts by the denominator. Extra ducts are provided in all
conduit lines where practicable, to provide for future growth
without tearing up pavements. The percentage of extra ducts
will vary for different locations, depending upon the estimates
of future requirements.
A typical feeder tap is illustrated in Fig. 8, which shows the
lead-covered cable rising up the pole and passing into the switch
box. From this point weatherproof cable is used, passing up-
ward and being connected to the feeder span which replaces the
galvanized strand wire used on other spans. An overhead feeder
tap is also shown on the same drawing.
The calculation of feeders may be conveniently based upon
two theorems in addition to Ohm's law :
1. The maximum drop, measured from one end of a uni-
formly loaded conductor, is one-half the drop produced by an
equal load concentrated at 'the distant end. Or stated in an-
other way. the maximum drop of a uniformly loaded conductor
is equal to the drop produced b}- the total load concentrated at
the center of the conductor.
2. This theorem has to do with the mathematical similarity
between moments in a mechanical system and drops in an elec-
trical system. For example, a beam supported at both ends and
having on it a certain distribution of load will be in equilibrium
when the sum of the moments about any point on the beam
is zero. Similarl}-, if a conductor has current fed into it from
both ends to sujiply any distribution oi lead upon it. there will
120
THE ARMOUR ENGINEER
[Vol. 4, No. 1
be for every distribution some point of division on the conductor
through which no current flows. This is the point of maximum
ch-op, and of equal drops from both ends, and the system may
then be said to be in equilibrium. In the electrical system we
have current, resistance, and drop, corresponding respectively
with load, distance and moment of the mechanical system. If
the conductor is of uniform size, we may use length of conductor
instead of resistance.
In making the detailed calculations for the various trolley
sections, two types of sections are to be distinguished:
TVP>1CA1_ OVEf?HCA,D
Fig. 8. Typical Feeder Tape
A. Isolated ; those receiving power from one station only.
B. Tie; those receiving power from two stations.
The fundamental assumptions underlying all these calcula-
tions are :
1. All stations operating at the same voltage. If this is not
the case, a simple modification may be made in the calculation to
properly divide for the difiference.
2. The load on each trolley section is uniformly distributed,
and feeder taps at approximately equal intervals reduce the load
uniformly. This is a condition which is approximately true in
city systems operating on a short headway.
3. No account is taken of the conductivity of the trolley
wires which are in parallel with the feeder for a portion of its
leaer :>ystems lor
Railways.
Rice.
120
THE ARMOUR ENGINEER
[Vol. 4, No. 1
be for every distribution some point of division on the conductor
through which no current flows. This is the point of maximum
drop, and of equal drops from both ends, and the system may
then be said to be in equihbrium. In the electrical system we
have current, resistance, and drop, corresponding respectively
with load, distance and moment of the mechanical system. If
the conductor is of uniform size, we may use length of conductor
instead of resistance.
In making the detailed calculations for the various trolley
sections, two types of sections are to be distinguished:
TYPlCAl- UMOtf?GROUNO
TVP»1C.>M_ OVE.t?HtA.O
Fig. 8. Typical Feeder Taps.
A. Isolated; those receiving power from one station only.
B. Tie; those receiving power from two stations.
The fundamental assumptions underlying all these calcula-
tions are :
1. All stations operating at the same voltage. If this is not
the case, a simple modification may be made in the calculation to
properly divide for the difference.
2. The load on each trolley section is uniformly distributed,
and feeder taps at approximately equal intervals reduce the load
uniformly. This is a condition which is approximately true in
city systems operating on a short headway.
3. No account is taken of the conductivity of the trolley
wires which are in i)arallel with the feeder for a portion of its
The Armour Engineer,
IV— 1, January, 1912.
Low Tension Feeder Systems for
Street Railways.
R. H, Rice.
Fig. 3. Hpol Mh|).
1, January, 1912.
ision Feeder Systems for
Street Railways.
R. H. Rice.
Sub ■ST-jvTioN
The Armour Engineer,
IV— 1, January, 1912.
Low Tension Feeder Systems for
Street Railways.
R. H. Rice.
Fig. 4. I-ond Uistribulidii Mai)' and 'Spidor Diagram.
ary, 1912.
der Systems for
ilways.
Rice.
::s
The Armour Engineer,
IV— 1, January, 1912.
Low Tension Feeder Systems for
Street Railways.
R. H. Rice.
Fig. 5. SlaMwn liond CenlerH.
i<eeder Systems for
:t Railways.
H. Rice.
III
The Armour Engineer,
IV— 1, January, 1912.
Low Tension Feeder Systems for
Street Railways.
R. H. Rice.
rig. B, I'eecler Diagr
rmour Engineer,
, January, 1912.
1 Feeder Systems for
;et Railways.
i. H. Rice.
■SOj-.-m OF=©3t?o ST
.RGEO DCTAIl_ NCAJ?
i-STATlON MO.I
The Armour Engineer,
IV— 1, January, 1912.
Low Tension Feeder Systems for
Street Railways.
R. H. Rice.
— °" •- -i-i^^
aUO-ST/kT.ON MO
1 •
3
,"
"
"
*
■*
.
"
.
-s-
= 56
3
f|o
5uo-ST/\Tiors rsoe
Fig. 7. Undergrouud Conduii-line diagram.
Engineer,
iry, 1912.
ler Systems for
ilways.
*ice.
RVES.
wheu the dis-
ircular mils are
ar mils when the
drop are given,
pee when the cur-
r mils are given.
.ES.
,a 500,000 circular-
png and carrying
i,000-foot ordinate
lil line, then hori-
pere ordinate is
;ion at 47.5 gives
-mils to carry 800
[th 30 volts loss.
e ordinate up to
horizontally until
i is crossed; the
section gives the
ar-mils.
f a 1,500,000 circu-
rry 600 amperes
Follow the 600-
ihe 40-volt curve,
be 1,500,000 circu-
late through this
of 9,350 feet.
The Armour Engineer,
IV— 1, January, 1912.
Low Tension Feeder Systems for
Street Railways.
R. H. Rice.
USE OK CUKVES.
flutl" the <lro|) when the dis-
urreiit anil circular mils are
To 1
the illstauee '
and
liar mils
EXAMPLES
1. Find the drop on a 500,000 clreular-
nill cable. 5,000 feet long and carrvlnR
4.i0 amperes. Follow R.OOO-foot ordinate
up to 500,000 circular-nill line, then hori-
zontally ■■ —
the volts drop.
2. Find the circular-mils to carry 80(1
amperes 8,000 feet with .SO volts loss.
Follow the 800-ampere ordinate up to
the 30-volt line, then horizontally until
the 8,000-foot ordinate la crossed; the
location of this intersection gives the
size as 2,250,000 elrcular-mlls.
" " ' 500,000 cir
Find the distance i
pere ordinate to the 40-voIt curve,
n horizontally to the 1,500,000 circn-
•rail line; the ordinate through this
he distance of 0,350 feet.
Fig. 9. Auxiliary Cu
Jan, 1912] RICE: LOW TENSION FEEDER SYSTEMS
121
length. In any ordinary case, the trolley wires form such a
small part of the total copper required for a section that their
neglect introduces small variation in the results of calculation.
Most, or perhaps all, of the above assumptions would be
modified in calculations involving an interurban road or a small
city system.
In the case of an isolated section, such as shown in M, Fig.
10, the calculation is very simple. The section AB of length L
.h
ICAL TROLCeV SECTIOMS
t=5^^
m
u,-fu
o"
E,= r(D,.-^)i.
--^n
o.
i
J
a
1
1
'^
1
J
1= c,*c,»c.
6
<
-
Fig.
1
).
Typical
Trolley Sections.
feet has a uniformly distributed load of / amperes, which is con-
sidered concentrated at the center of the section. If r is the
resistance per foot of feeder, the drop from the power station
to the nearer end of the section is r D I. The added drop to
the end of the section is ^ r L /, if the feeder is continued un-
diminished in size to the end of the section. The total drop is
as given by the equation
£ = r (D+/.Z.)/. (1)
122 THE ARMOUR ENGINEER !Vol. 4. No. 1
If the feeder is reduced in size as the load decreases, die maxi-
mum drop at the end of the section Avill he somewhat greater than
that given.
The calculation of a tie section is a little more complex.
Take, for instance, the simplest case, illustrated by N in Fig. 10,
in which the main feeder betweejn stations is assumed to extend
the entire length of the section and to be uniform in size. The
section A B has a uniformly distributed load of total value /
amperes, of which /, amperes are assumed to come from station
5"! and I. amperes from 5... P is the point of division of load
beiween the stations, and is the point of maximum drop on the
section. We are usually concerned in knowing the load on each
station and the maximum drop on the section.
To determine the load /, on station 5",, take moments about
5o. These moments must be so chosen that they will involve
only the one unknown whose value is sought, otherwise the
solution of simultaneous e(|uations becomes necessary, and much
needless labor is introduced. Tn this case assume the total load
concentrated at the center of the section; then its moments about
So would be
CD, + /2Z.)/.
The moment of the load /, at S,. which would just balance
this moment about S.,, is
Kquating and solving, we get the load on 5", equal to
which shows that the load on one station is equal to the total
section load, multiplied by a fraction whose numerator is the
distance from the second station to the center of the trolley sec-
tion, and whose denominator is the distance between the stations.
The location of the ])oint of division of load is readily de-
termined. Since the load is uniformly distributed, we get
L^ = ^jL (3)
■ The maximum drop occurs at P and is
A second ty])c of tic secli(^;i is one in which tlie main feeder
between station's does not ])arallel the trolley sectifMi throughout
Jan., 1912] RICE: LOW TENSION EEEDER SYSTEMS 123
its length, as shown at O, in Fig. 10. The loads on the three parts
of 'the irolley section arc C,. C,, and C... We then have
/ = C, + C + C. (5)
I'o find the load on 5^1 we take moments as before about 5"o,
which gives
C, D, + C, (D, + y2U) + Q (D, + L,) = /, (D, + L, + D,).
Multiplying out and factoring, we get
(C, + C + Q ) D, + C, y^L, + CM, = I, (D, + L, + D,),
which, by use of (5) reduces to
ID, + C, yi, + r, L, = /, (D, -\- L, + D,),
and hence
The load distributed over the distance A' is /, — C... and since
the distribution is uuifnrm. we have
X = ^L. (7)
The drop from 5", ti> tlie point of division P is then
£j = (/,_C) K^r.r + /,!),;- (8)
which is the maximum drop on the feeder, but which may be
exceeded at the ends A and B of the trolley section.
As an aid in making calculations in these and similar cases,
there was prepared a series of curves as shown in Fig. 9. The
radial lines represent the relation between lengths and resistances
for various sizes of cables. The other set are curves of equal
drops, and show the product of various resistances and currents.
There are five elements entering into the curves : circular-mils,
current, drop, distance and resistance, and the curves are so re-
lated that the desired quantity may be read directly from the
chart when the other values are known.
The method as given here has been found to be very satis-
factory in practice. The results are obtained without any "cut
and try" process, and by a little familiarity with the method, it
is possible to work quite rapidly. The final results show those
facts which it is usually necessary to know, viz., the division of
load between stations and the maximum drop on the cable.
C. B. & Q. R. R. TRACK ELEVATION WATERPROOFING.
BY G. A. HAGGANDER.*
One of the hardest problems presented in connection with
the concrete track elevation subways on the C. B. & O. R. R. in
Chicago is the waterproofing. These subways are constructed
with concrete slabs or girders seven feet wide, built away from
the work, resting on concrete abutments at the street lines and on
concrete columns and cross girders at the curb lines and center
line of street. There is a joint between the slabs longitudinally
across the bridge, also one over each cross girder and one at the
bridge seat on top of the abutment. The slabs over the street are
2'9'' thick over the street girder and 2'6" thick over the curb cross
girder The slabs over the sidewalk are I'S^" thick over the curb
cross girder and 1' 2>^" thick over the bridge seat.
The first waterproofing was done in 1906. The most natural
method seemed to be the closing of the joints. In order to do
this the first slabs were built with the sides ofifset as shown in
Fig. 1. The open space between them after they were set was _^4"
at the lower part and 1^4'' at the upper part. A board was put
under the joint where the bridge seats did not prevent the mortar
from leaking through and the J4" space was filled with cement
grout to within 1^'' of the offset. The next 3'' was calked with
oakum soaked with R. I. W. waterproofing compound. The rest
of the opening was filled with cement grout and over the entire
joint were placed three layers of felt painted with R. I. W. com-
pound.
This method was not effective, the subways leaking very
badly. It did not prevent the water behind the abutment from
running through the mortar joint down the face, and the joints
cracked open, letting the water through. It was thought best to
fill the joints with some elastic material and six schemes were
tried as shown in Figs. 2 to 7 inclusive. They all consist of dif-
ferent arrangements of oakum, cement grout, tar or some asphalt-
ic compound and sand. In every case the bottom of the crack
was calked with oakum to prevent having to use boards under
the joints. None of the methods were even partly successful
until the sixth was tried. This one failed when the tar leaked out
through the oakum, but this was prevented to a great extent by
"Class of 1007. Office Euiiineer, Brulfie Dep.irtment, C. B. & Q. R. R., Chicajro,
Illinois.
Jan., 1912J HAGGANDER: C. B. & Q. R.R. WATERPROOFING 125
3
:ff
=T^
i
"^ t^i^T
S
j°i I
^ f
126 THE ARMOUR ENGINEER [Vol. 4, No. 1
the use of sand above the oakum. At first the joints were water
tight but the jar and shght working about of the slabs loosened
the oakum, letting the sand and tar escape. The joint at the abut-
ment also leaked very badly. In addition to this the water found
its way through two of the slabs themselves.
These methods had been tried during 1906 and 1907 and in
1908 it was thought best to waterproof the whole surface of the
bridge and to aid the water in running off by a system of drain
pipes.
The method shown on the right-hand side of Fig; 8 was the
one adopted. The slabs were built with a Yi" batter on the sides
instead of having an offset so that they could be calked from
above. After calking the cracks with oakum they were filled
with cement grout. The vertical offset of 12>4" over the curb
cross girder was rounded off by a fillet of concrete. The abut-
ment back of the bridge seat was built up to the level of the top
of ihe sidewalk slab. The lifting stirrups or hook bolts by means
of which the slabs were placed were cut off at the surface of the
concrete and the surface of the bridge swept clean. As a rule
only one track or a width of fourteen feet could be waterproofed
at one time on account of operating conditions and the force used
was of such size that this was one day's work. A strip about
four feet wide was mopped with a coat of asphalt which had
been heated to the melting point. Four kinds of asphalt were
used at different times, Sarco No. 6 and No. 651, melting point
160°. Barber Asphalt Co.'s positive deal "A," melting point 140^,
Texaco and Warren Chemical Co.'s asphalt cement. On this was
])ut a strip of eight-ounce open-mesh, first quality bu-rlap. 42"
wide. This came in 2,000 yard bundles and was made into rolls
coiled on a VA" gas pipe seven feet long. It was applied by
starting at one end of the bridge and rolling it across. A man
followed the roll and swept out any wrinkles so as to give a
smooth surface and bond it to the asphalt below. The top of
this burlap and a strip of concrete about fifteen inches wide to
one side of it was then painted with hot asphalt and another layer
of burlap laid, covering two-thirds of the first layer, the rest
lapping over on the concrete. This burlap and the adjacent con-
crete were then painted and another strip laid covering one-third
of the first and two-thirds of the second strip. In this way the
burlap was made three-ply. Care had been taken that the tem-
perature of the asphalt was not high enough to burn the burlap.
A melting point of 160° Fahrenheit gave the best results, as
asphalt with a higher melting point burns the burlap and becomes
more brittle at low temperatures. After three ply of burlap had
Jan., 1912] HAGGAXDl-.K: C. I'.. <S; O. R.R. \VATERPR()( )F1 XG 12,'
128 THE ARMOUR ENGINEER [Vol. 4, No. 1
been laid the whole surface was again mopped with asphalt and
a protecting layer of mastic put on. This layer consisted of one
part of asphalt and four parts of dry engine sand. The asphalt
was heated to the melting point in one kettle and the engine sand
heated over an iron plate. They were mixed and stirred in
another kettle until of the right consistency, a slow fire under the
kettle keeping the mixture plastic. This mastic was dipped into
iron wheelbarrows, wheeled on broad runways laid on the bur-
lap, dumped, and finished with wooden trowels to a thickness of
one inch. The top of this mastic was mopped with asphalt. This
one-inch layer was not intended to aid in waterproofing, but was
put on to protect the burlap from the ballast and track tools.
The drainage system consisted of inverted eight-inch half-
tiles laid between each track leading from the hump over the curb
cross girder to the back of the abutment. Back of the abutment
and about one foot below the bridge seat was placed a board on
which rested another eight-inch half-tile. The end of the burlap
rested in this and another four-inch tile was laid in it. The tiling
was given a slope of about 1% toward down-pipes. These down-
pipes were fifty feet apart and ran down the back of the abut-
ment to the sidewalk level, then through it and under the side-
walk to the gutter. The mastic was protected by ^^" of roofing
gravel to keep the sharp edges of the crushed stone ballast from
cutting into it. The crushed stone ranged from ^" to 2j^" and
was used to allow free movement of the water through it.
It was not long, however, before the boards and tiling
crushed down into the newly made fill, giving no means of es-
cape for the water which ran off the bridge, causing it to back
up above the level of the bridge seat and find its way down the
face of the abutment. This was prevented by building the con-
crete so as to over-hang the back of the abutment about one foot
and letting it run down the back about the same distance. The
top of the abutment was painted with a heavy coat of tar paint
before setting the slabs. This effectually sealed the joint. In
all abutments built after this experience a notch was left in the
top, as shown on the left hand side of Fig. 8, in which to place
the tile and prevent its crushing down.
After the first year's experience with this waterproofing, it
was thought advisable to increase the thickness of burlap to
five-ply. Many leaks had developed over the curb cross girder
due to the tearing of the burlap over the hump at the cross
girder joint. This seemed to be caused by a slight expansion and
contraction of the slabs and tendency of the waterproofing to
travel down the slope of the hump due to the load of an engine
Jan., 1912] HAGGANDER: C. B. & Q. R.R. WATERPROOFING 129
above it. The crack which opened up was sometimes as wide as
ji". Some method of allowing for this had to be used so a one-
inch pipe was laid over the joint, the burlap put on in the regular
way, the pipe withdrawn, and the mastic applied. This scheme
was used on all work during the year 1909 and thereafter.
Another precaution was taken by puttiing a flatter slope on the
hump, the concrete being run out about 3' 6" so that the tendency
of^ the waterproofing to slide down was diminished. The
bridges waterproofed during 1909 are almost devoid of leaks.
The bridges which were waterproofed by filling the joints,
in 1906 and 1907, have since been waterproofed bv the latter
method. Those which were waterproofed during' 1908 leak
where the waterproofing has cracked and repair work was started
during the fall of 1911. In this repair work the mastic is chipped
off for a distance of eight inches each side of the crack. The
burlap is cut awav from the crack to the edge of the mastic on
the street slab. A one-inch pipe is then laid over the joint and
the mastic and the burlap on the street slab lifted up so that a
new strip oi burlap can be inserted under it. It is lapped under
about nine inches and then laid over the old burlap on the slope.
Five-play are laid, each painted with asphalt, the pipe withdrawn
(thus allowing for expansion), and new mastic put on.
Records have been kept of the dates when different portions
of the bridges are waterproofed. This repair work has shown
that during the cold weather the asphalt that is applied to the
cold concrete becomes chilled and does not penetrate the burlap
which is laid on it. If the weather is very cold, even the coat
which is mopped on this first layer of burlap does not penetrate
it due to the fact that the cold concrete has chilled it. The sec-
ond layer of burlap seems to have been pretty well saturated in
all cases, while the third layer is thoroughly saturated. In water-
proofing which was laid during warm weather all of the burlap
is in a good state of preservation, having been well saturated with
the asphalt. The mastic seems to protect the burlap very well,
although in some cases track tools have penetrated it, causing
leaks. The smaller sized crushed stone of about ^" does not
penetrate the mastic as much as the larger sizes on account of
the more uniform bearing.
The average force used on this work consisted of:
1 Foreman, @ $ .33 per hour
3 Finishers, @ .25 " "
1 Kettleman, @ .27i/, "
8 Laborers, @ .20 " "
130 THE ARMOUR ENGINEER [Vol. 4, No. 1
These men waterproofed a bridge 14 feet wide and 75 feet
long, or 1,050 square feet per ten-hour day.
The average cost of the later waterproofing was a httle less
than fourteen cents per square foot for the burlap and the mastic,
not counting the concreting over the hump and bridge seat and
the filling of the joints with mortar. This work was done by
another gang the day previous to the waterproofing. It adds
about three cents per square foot to the cost, making a total of
seventeen cents.
This work was done under the direction of C. H. Cartlidge.
Bridge Engineer, and L. J. Hotchkiss, Asst. Bridge Engineer, and
was directly in charge of F. H. Cramer, by whom much of the
material in this article was furnished.
THE LIFE OF AN AUTOMATIC SPRINKLERt
BY C. R. ALLING.*
The life of an automatic sprinkler is not indefinite under any
conditions, and under certain conditions it may be comparatively
short. The sprinkler fire record would 'be even more favorable
than at present were this more generally known and proper pre-
cautions taken to protect sprinklers from influences detrimental to
their operation. The automatic sprinkler is generally recognized
as the most effectual fire extinguishing agent yet developed, and
considering all that it has to contend with in the way of faults in
design, errors in construction, mistakes in installation and mis-
treatment in service, it is surprising that its general showing is so
favorable. It is surely deserving of better supervision, care and
maintenance.
The object of this article is to point out a few of the more
common causes of failure of automatic sprinklers which have
come under the attention of the engineers at the Underwriters'
Laboratories, in the hope that persons interested may co-operate
with the Laboratories toward an advance in the state of the
sprinkler art, particularly that portion of sprinkler practice for
which building owners and occupants are chiefly responsible.
Up to the present time no sprinkler has been designed which
will prove entirely reliable when exposed to very severe loading
or corrosive conditions for long periods of time. The life of
some sprinklers is longer than others, due to their more perfect
design or excellence in manufacture, thus increasing their ability
to better withstand these influences. Also, the life of some types
of sprinklers will be longer where there are no corrosive atmos-
pheric conditions than other types for the same reasons. Thus it
will be seen that the life of a sprinkler depends on the excellence
of its design and construction, as well as upon the influence of
surrounding conditions.
Some types of sprinklers were, in the early days of manufac-
ture, made of cast iron. These were unduly susceptible to influ-
ences of corrosion, and experience soon taught the pioneers of
this industry that most all parts of a device of this character
should be made of metals which would be least subject to deterio-
*Class of 1907. Assistant Eugineer, Underwriters' Laboratories, Chica.£?o.
■j-A paper publislied in the Quarterly of The National Fire Protection As-
132 THE ARMOUR ENGINEER [Vol. 4, No. 1
ration and, therefore, for the last twenty-five years, practically
every manufacturer has employed brass, bronze, or some similar
alloy, which will be the least affected by conditions to which^ an
automatic sprinkler is normally subjected. At the present time
there is probably not a sprinkler in service in which the more es-
sential parts are made of any material other than bronze or some
other copper-tin alloy.
Tests conducted at the Laboratories upon old sprinklers taken
from the field, and having been in service for varying lengths of
time, show that the greatest and the most common causes of fail-
ure are corrosion and loading. Corrosion usually acts upon the
solder, changing the chemical composition upon the surface and
forming a crust. Continued exposure will sometimes entirely dis-
integrate the solder and weaken the members. This is especially
so in the early types of sprinklers. A good example of how con-
tinued neglect will result in an absolute failure of the sprinkler
can be seen in Fig. 1. This sprinkler had been in service about
ten years over a dipping tank in a plating establishment, and ap-
parently no means of protecting it from the surrounding condi-
tions had been provided. In this sample the solder was entirely
eaten away, the parts being held together bv the corrosion, and
the sprinkler being absolutely inoperative. Even if the releasing
device had operated during a fire the distribution would have been
inadequate, as the deflector had been destroyed by corrosion.
A large number of sprinklers have been tested at the Labora-
tories from equipments where the corrosive conditions are fairly
severe, such as paper mills, stables, fertilizer plants, tanneries, dye
houses, sulphite mills, aniline works, soap factories, etc. These
tests indicate that no sprinkler will remain in good condition in-
definitely and that the only way in which it can be determined
whether the heads are in a reliable and operative condition in-
test representative samples under conditions that may be met with
at time of fire. It has been found in a number of instances that
although the sprinklers appeared to be in fairly good condition
before the test, they were absolutely inoperative, while in other
cases the corrosion appeared to be sufficient to render the heads
inoperative, but upon test they were found to be reliable.
Various attempts have been made by the sprinkler manufac-
turers to protect heads which were to be installed in locations
unusually subject to corrosive influences, but none of these have
been entirely successful. One method consisted in covering the
entire sprinkler above the wrench head with a glass cap, the idea
being that the heat from a fire would either be sufficient to crack
the glass or else the heat would be transmitted through the glass
Jan., 1912] ALLING: AUTOMATIC SPRINKLERS 133
to the soldered joint so that when the releasing device operated
the water pressure would be sufficient to blow off the cap. The
principal objection to this was that the sensitiveness of the sprink-
ler was materially decreased.
Another method which is probably more successful consists
of coating all parts of the sprinkler with mineral wax composi-
tions having melting points below that of the ordinary degree sol-
der, so that in case of fire the composition will melt and leave the
Fig. 1. Sprinkler Inoperative. Solder Entirely Eaten Away; Parts Held to-
gether by the Corrosion.
parts free from corrosion before the heat becomes sufficient to
fuse the solder. This method of protecting the sprinklers has
generally given satisfactory results for a limited time, the princi-
pal defects being that under severe conditions the coating dries
out or becomes cracked by expansion and contraction, thus allow-
ing the corrosion to work in and attack the metal parts. No suc-
cessful method of protecting high degree sprinklers has yet been
devised.
134 THE ARMOUR ENGINEER [Vol. 4, No. 1
Sprinklers located in plants where they are subject to loading
will become inoperative in a comparatively short time, the clog-
ging effect of the loading being often just as effective in rendering
the sprinkler inoperative as severe corrosive influences.. This is
evidenced by tests of sprinklers taken from various mills where
they are subjected to loading from paper pulp, sawdust, drying
oils, lint, caked dust from grinding processes, etc., etc. The char-
acter of the loading is somewhat different in almost all cases,
^^
\^
m^
•' />
Fig. 2. Sprinkler Inoiierative. Loaded and Clogged with Paper Pulp.
varying from a thin hard crust to a thick, heavy, fibrous coating.
All of these loadings tend to accomplish the same result ; i. c, to
clog the parts so that the releasing device cannot operate freely,
if at all. Fig. 2 shows the condition of a sprinkler received at the
Laboratories for test. This head was taken from a paper mill
and was entirely covered with paper pulp.
Probably one of the most frequent objections that an inspec-
tor makes in looking over an equipment is to the coating of the
Jan., 1912] ALLING: AUTOMATIC SPRINKLERS
135
heads with paint or calcimine. In calcimining or painting a ceil-
ing there is great danger of the sprinklers becoming slightly
coated; and if this should occur at the soldered joint of the re-
easing device, the sprinker, especially if it is one of the older types,
is very liable to be in a questionable condition. Its appearance
may not indicate its true worth, and its reliability can only be de-
termined by suitable tests.
Many of the equipments at present in service are provided
Fig. 3. Sprinkler Inoperative at 100-pound Pressure.
with sprinklers which at the time of installation were probably
as good as could be obtained, but do not now come up to the pres-
ent state of the sprinkler art. These heads, even when new, were
not always entirely reliable, and the years that they have been in
service have accentuated their defects. For example, one type of
sprinkler that was manufactured about 1898 was so designed that
the levers of the releasing device are nearly on dead center, the
result being that when the solder on the link becomes oxidized or
slightly corroded or loaded, there is not sufficient motive power
136 THE ARMOUR ENGINEER [Vol. 4, No. 1
to operate the parts of the releasing device. Fig. 3 shows one of
these sprinklers, which was inoperative at a pressure of 100 pounds
even after the fusible link had been removed.
On the other hand, other types of sprinklers have been manu-
factured in which the off-center distance of the levers was great,
resulting in an abnormal strain on the soldered joint, the head
finally opening prematurely. It is essential that the sprinkler be
so designed that it will safely withstand the stress to which it is
normally subjected, and at the same time insure a sharp reliable
release of the parts under all reasonable conditions of service.
These features have been embodied to a greater or less extent in
all the more modern sprinklers.
The life of an automatic sprinkler depends primarily upon
its design, accuracy in workmanship, and ability to retain sufficieni
motive power, independent of water pressures, to overcome all
reasonable obstruction to its free operation for extended periods.
Tests made at the Laboratories upon old sprinklers indicate that
inaccuracy in the machined surfaces and bearings are largely re-
sponsible for the slow actions of the releasing device, resulting in
partial opening of the link, and in some cases the resealing of the
solder by the water. This is caused by a slight set between the
parts, and consequent loss of the motive power necessary for the
proper operation of the parts. The more modern sprinklers con-
tain features which increase the chances of the releasing device
operating favorably, or aid in its continuous movement. In some
cases the frames and levers are so designed that when the desired
load is placed on the link or strut, only a portion of the elasticity
is utilized, thus retaining the necessary power for the free and
reliable action of the releasing device when the fusible element
operates. In other cases, springs, diaphragms, etc., have been in-
corporated to insure the proper operation of the parts. In one
of the earlier types of sprinklers, which was approved quite gen-
erally, the design was such that after its operation the link could
be replaced in the sprinkler as installed in the piping. This per-
mitted a stock of fusible links to be kept on hand for replacements,
when necessary, which led to misuse in the field ; and when this
stock was exhausted, or not easily found, the two arms were fre-
quently wired together and, as a result, the sprinklers were entirely
inoperative.
Another feature that was not entirely realized in the early
types was the freedom of release of the caps and discs from the
seats. A large number of sprinklers are in service at the present
time which are open to these most important defects, and the only
way in which these may be detected is by laboratory tests. In
some sprinklers lead and other soft metal discs were formerly
Jan., 1912] ALLING: AUTOMATIC SPRINKLERS 137
employed, and after having been in service for a short time it was
found that the seat ring gradually became embedded to such an
extent that the sprinkler was rendered inoperative, the cap remain-
ing on its seat even under very high water pressures. Cases are
on record where pressures in excess of 150 pounds were neces-
sary to release the cap. Recent tests of another type of sprinkler
show that the caps and block tin discs have become wedged in the
discharge orifice, so that pressures in excess of 100 pounds are
required to release them after the link and lever portions of the
releasing device have operated. Also in this same sprinkler, the
caps have released satisfactorily, but the perforated block tin discs
adhered to their seats, requiring velocities produced by pressures
of 150 pounds to force them ofif. The opening through these discs
is three-eighths of an inch, and as a result the water discharged is
reduced about fifty per cent under normal conditions.
From the foregoing it would appear that property owners, as
a rule, need to be brought to a realization of the fact that the life
of an automatic sprinkler is »ot indefinite. The need for replac-
ing old, obsolete and injured heads should be constantly borne in
mind. The exercise of care and intelligence in making such re-
placements is, of course, essential. That more and more attention
is being paid to this important matter is evidenced by the con-
stantly increasing volume of examination and test work made at
the Underwriters' Laboratories on specimen heads sent in from
equipments in the field. Thorough tests, such as are conducted at
the Laboratories, are perhaps the only true criterion on which
action should be based, in the majority of cases, and as these
tests are made promptly and without charge there would seem to
be no justifiable reason for depending on field tests or less thor-
ough analytical work than the Laboratories conduct. Automatic
sprinklers are, of course, sometimes found in service which will
manifestly fail under ordinary fire conditions, but in the majority
of cases, the demarcation between reliability and unreliability in
the device cannot be accurately made except by Laboratory tests
conducted bv men trained in the business.
NATURAL GAS AND ITS DISTRIBUTION.
BY J. G. HATMAN.*
The natural gas found in the gas wells of southern Kansas
and Oklahoma has a specific gravity of about 0.6 and a heat value
of 950 British thermal units. This gas has a chemical formula
of CH^, is odorless, and contains practically no moisture. To
burn properly requires a mixture of air to gas of about eight or
nine to one. The gas is used in the power plants of the packing
and street railway companies and various other industries for
boiler fuel. The cost of power is about the same using coal or
gas. A three hundred horsepower boiler is usually equipped with
twelve burners, making good control possible. Each burner will
consume about twelve hundred cubic feet of gas per hour at a
pressure of one-fourth pound. The fact that these plants can
obtain the gas only for a period of five months during the sum-
mer, due to a lack of pressure in the winter, necessitates a change
of stokers twice a year at considerable expense. This expense is
partially overcome by the cleanliness of gas-burning, and no coal-
and ash-handhng and less labor. The gas is used by domestic con-
sumers in hot air furnaces, for cooking and lighting, so that coal
is little used, making a clean city.
The fields are gradually giving out and new fields are opened
further south so that nearly all the gas used in eastern Kansas and
western Missouri now comes from Oklahoma. The main com-
pressor station of the Kansas Natural Gas Company is in the
southern part of Kansas ; this company leases the wells and sells
the gas to local companies in various cities, such as St. Joseph and
Kansas City, Missouri, and Kansas City, Leavenworth and To-
peka, Kansas.
From this station the gas passes through three more pumping
stations before it reaches Kansas City. At the southern limits of
the county is a large reducing station through which all the gas
used by both Kansas Cities passes, amounting to about fifty to
seventy million cubic feet of gas per twenty-four hours.
By referring to the accompanying map it will be seen that
Kansas City, Kansas, is divided into three districts, Kansas City,
Rosedale, and Argentine. Station "A" is the reducing station
above mentioned. Here the gas is reduced from a pressure of 175
*Class of 1010. Assistant Superintendent, Wyandotte County Gas Co., Kansas
City, Kansas.
GOV£ffNOR 3THT/ON3
fTND
DISTRI0UT/ON /^HINS
//V
Sa>/e I"' I Mile
1 i
140
THE ARMOUR ENGINEER
[Vol. 4, No. 1
pounds or 200 pounds to 50 pounds. From this station the gas
passes through two sixteen-inch mains to the station "C," where
it is again reduced in pressure to twenty pounds . This is the dis-
tributing station for Kansas City, Argentine, and the business dis-
trict of Kansas City, Alissouri. At points along the sixteen-inch
main, shown by "x," small two-inch mains are taken through
underground reducing stations. These stations are about eight
feet by ten feet by three feet, and in each are two regulators hav-
ing an inlet of three inches. These regulators reduce the pressure
Figr. 1. Typical High Pressure Service.
to fifteen pounds. Each of these stations supplies about one hun-
dred consumers.
At station "B" a ten-inch main is taken from the large main
and passed through a ten-inch regulator which reduces the pres-
sure to ten inches of water and supplies that part of Rosedale
shown within the ten-inch line.
From 'the station "C" a six-inch main carries the gas to Ar-
gentine at a pressure of twenty pounds. This six-inch main is laid
as nearly as possible through the center of Argentine, and from
this three-inch, two and one-half, and two-inch mains are taken.
All the mains and each service up to the meter in every consum-
er's house in this district carry a pressure of twenty pounds. A
typical high pressure service is shown in Fig. 1.
Jan., 1912] H ATM AN: NATURAL GAS DISTRIBUTION
141
Before the meter a governor is set, shown in Figs. 2 and 3.
This governor consists of a rubber diaphragm set between xwo
metal plates. The desired outlet pressure, usually five to eight
inches of water, is set by means of the pressure screw. In the
mercury seal two'pounds of mercury are placed. If a diaphragm or
any other part of the regulator should break, as soon as the pres-
sure reached thirteen inches of water the gas would go through
the mercury seal into the atmosphere through the vent pipe; this
vent pipe is set against the outside wall of the building and the end
terror* ^utr-^ o^
f«vo
^^
rig. 3. Construction of Higli Pressure House Governor.
External View of House Governor.
is about eight or ten feet above the ground. This governor is
used in two sizes, depending upon the consumption. An inlet and
outlet of three-fourths of an inch will pass two hundred cubic feet
per hour, and an inlet of three-fourths of an inch with an outlet
of one inch will pass four hundred cubic feet per hour with a
pressure of ten inches of water.
From the station "C" a twelve-inch main carries at a pressure
of twenty pounds the gas to Kansas City, Kansas, to eight district
stations. At points just outside these stations, branches are taken
142
THE ARMOUR ENGINEER
[Vol. 4, No. 1
off and the gas passed through eight-inch or ten-inch governors
as shown in Fig. 4. These stations are set in parallel so that any
one station can be cut out at any time without affecting the pres-
sure in the district cut out. There are two of these governors in
each of these stations, each reducing the pressure from twenty
Reynolds Gas Regulator Co., Anderson, Ind.
Fig. 4. Type of Regulator Ised in District Governor Station.
pounds to ten inches. Between the governors on the outlet side,
and above them, a cylindrical tank is built about four feet in
diameter and three 'feet high, which acts as a pressure seal similar
to the mercury seal previously described. This tank is filled to a
depth of about twenty inches with gas oil, which has a specific
gravity of about eight-tenths, causing the^^tation to blow through
a metal stack in the roof when the ou'tlet pressure reaches sixteen
Jan., 1912] HATMAN: NATURAL GAS DISTRIBUTION 143
inches of water. These stations are built of red brick, have corru-
gated roofs, and are about fifteen feet by twenty feet by eight feet.
In the northwest section of the city, suppHed by station num-
ber 7. all the services have a low pressure regulator. This regu-
lator has an inlet pressure of sixteen inches and an outlet pressure
of eight inches of water. Two eight-inch district valves can be
closed and the station raised to sixteen inches, which makes this
district wholly controlled by this station. In the warm weather the
valves are opened and all stations set at an equal pressure of ten
inches of water.
At station number 5. but entirely independent of the station,
is a holder tank having a capacitv of seven huncTred and fifty thou-
sand cubic feet, which is kept filled^'for emergency cases, such as
breaks in the main line before it reaches the city, and also for
cold weather. In cold weather the pressure at the station ''A" gets
as low as twenty or thirty pounds, which makes the low-pressure
side of the governor station get as low as three or four inches of
water. When this happens the reserve supply is pumped into
either the high-pressure or low-pressure side. This is usually
done at 'the time of the peak load, between five and eight o'clock
at night. During tlie night, the gas in the high pressure mains
packs up, due to the decrease in consumption, at which time the
holder is filled for the next peak load. This decrease in pressure
occurs whenever the weather gets cold, which is about twenty de-
grees above zero for this section of the country.
From each of the eight district stations, either a ten-inch or a
twelve-inch cast iron main is laid through the center of a district,
and this main is designated as a trunk main. From this main,
four-inch, and six-inch mains are taken at intersecting streets.
The size of the service pipe in low pressure districts is approxi-
mately gauged by the followng short table:
1 to 8 room house ^. 1^-inch pipe
8 to 12 room house 1^-inch pip^
Over 12 room house 2 -inch pipe
For small boilers three-inch pipe is used and a large power
plant installation is usually made with six-inch or eight-inch pipe.
All high-pressure mains are laid of wrought iron pipe wth screw
joints, while the low pressure mains are of cast iron pipe with
cement joints. Services are laid with either steel or wrought iron
pipe. ^";
THE CHICAGO RAILROAD TERMINAL PROBLEM.t
BY L. C. FRITCH.*
One of the most important problems which require solution
in Chicago at the present time is that of the railroad terminals.
Chicago is the largest railroad center in the world. Thirty or
more separate railway corporations have terminals here, represent-
ing over 85,000 miles of road, or over one-third the entire mileage
of the United States.
The traffic originating at Chicago, destined to Chicago, and
passing through Chicago is so vast in volume that it has long out-
grown the facilities required to promptly handle the maximum
volume expeditiously and economically. There is no problem
confronting Chicago^ that so vitally affects its commerce as the
railroad terminal question, but despite this fact no comprehensive
plan or investigation into the question has been undertaken. _ It is
of far greater importance to the commercial supremacy of Chicago
than tile question of docks and wharves, smoke elimination or
electrification of railroad terminals. In fact the three questions
mentioned depend for their proper solution entirely upon the
main problem and cannot be solved until the entire railroad situa-
tion has been revised upon a scientific basis and through the co-
operation of all the railway lines.
The lack of co-operation of the railways has been in the past
the principal cause of the failure to undertake this vastly im-
portant matter. The selfish interests of a few prevented the
accomplishments of a purpose which would result in the benefit of
all lines and the entire community. There has been, however, a
change in the sentiment of some of the important factors in this
matter and from recent developments it appears that some definite
results will be secured through the joint co-operation of civic
bodies and the railroads.
The solution of this whole question recjuires a comprehensive
investigation scientifically conducted with due regard to the inter-
ests of all transportation lines and the public and its efifect upon
the future growth and welfare of the community. Obviously,
this is absolutely necessary if a condition is to be avoided which
*Chief Engineer, Chicago Great Western Railroad, Chicago.
+A paper read before the Armour Institute of Technology Alumni Association
at the Mid-Winter Dinner, Hotel Sherman, Chicago, December 19, 1911.
Jan., 1912J FRITCH: CHICAGO R. R. TERMINALS 145
is being rapidly reached, and which will make impossible the car-
rying out of a plan already under way of making Chicago the
greatest commercial and civic center in the world.
A thorough treatment of the railroad terminal problem in Chi-
cago would require more time than is here permitted, but the
main features may be briefly reviewed. Classified by service, as
service regulates the use of the facilities, terminals may be grouped
into two main classes, passenger terminals and freight terminals.
Passenger terminals may be subdivided into (A) those for ex-
press or through passenger traffic, and (B) those for local or
suburban passenger traffic. Freight terminals may be subdivided
into (A) those for local traffic, and (B) those for through traffic.
Passenger Terminals.
The passenger terminals of a railway largely fix in the public
mind the character of the lines using same, by reason of the mora
intimate acquaintance with them than with freight terminals, and
yet the freight terminals have a greater bearing upon the pros-
perity and interests of a community than have the passenger
terminals. Passenger terminals are provided at greater relative
cost than are freight terminals, and produce relatively less re-
turns. As a rule, more attention is given to passenger facilities
and less to freight facilities than their relative importance de-
mands. This is aptly illustrated in our very midst by a most
recent example.
The present passenger terminals for express or through
service in Chicago are not adequate to the needs nor the impor-
tance of the second city in the country. There is not one pas-
senger terminal station in Chicago today that is located in the
proper place, nor adequate in its facilities to meet the require-
ments of the next fifteen to twenty years. One new station has
just been built and three new ones are contemplated. It is to
be regretted that the one just built should not have been deferred
until a comprehensive plan could be worked out to determine the
best location and its utility for the general purpose.
One Union Station for all lines entering Chicago would be
inadvisable, but one general location providing for a group of
stations would be practicable and in the interest of economy and
the public welfare. The general plan proposed by Mr. Delano,
locating all passenger stations in groups along Twelfth Street, is
the most practicable one yet suggested. The advantages of such
a plan are at once apparent in considering the economy of space,
joint use of facilities and convenience of access by the public.
The further advantage of permitting the necessary expansion of
the business district of Chicago is of the utmost importance. The
146 THE ARMOUR ENGINEER [Vol. 4, No. 1
effect upon the growth of the business district of Chicago if pas-
senger terminals are rebuilt in or near their present locations
cannot be other than detrimental and for this reason, if for no
other, the whole problem should be carefully considered before
expensive mistakes are made. Land values in the business dis-
trict of Chicago are too high to permit the use of one terminal
passenger station by one line and the policy should be that as
many lines should be permitted to use one station as can be con-
veniently accommodated therein with joint use of approaches
thereto. The convenience and value of all stations being located
at one general location is appreciated in the interchange of
through passenger trafific. One line occupying a station alone is
cut off from this valuable feature, and furthermore has the large
fixed charges and maintenance to bear alone, which would other-
wise be divided among many lines.
The future growth and prosperity of Chicago as a city, and
the interests of the railways, demand that no new passenger
terminal shall be built until every possible consideration has been
given to it from all viewpoints.
The local or suburban passenger terminal facilities are now
principally merged with through passenger facilities except in
a few instances. The future of these facilities will require that
they be separated entirely from through terminal stations, which
will be located farther from the business center. The steam rail-
way suburban traffic will in time either be abandoned, or electric
power will replace steam power. It is a well known fact that
steam power on trains of short runs, few cars, and frequent
operation is unprofitable as against electrically-operated service,
and where density of service does not warrant electrification of
suburban lines the traffic is not profitable and will be absorbed
by electric lines.
Freight Terminals.
The freight terminals of a railroad are the most important
from the standpoint of revenue to the railroad and the commerce
of the community. They are not so prominent, however, in the
public eye, and few really know much about their extent and their
actual location. So important, however, are these facilities, and
so widely scattered, that they occupy an area in Chicago about
equal to our business district. As a rule the facilities have not
kept pace with the growth of the traffic and at times of heavy
business serious congestions result. The main lines of our rail-
ways are highly developed, but terminals are in many instances
of the same capacity as in years past, with few additions.
Of first importance to a commercial community are the local
Jan.. 19121 I'RITCIl: CHICAGO R. K. TKR M 1 XA LS 147
freight terminal facilities. These must include freight houses,
platforms, etc.. for receiving and forwarding less than car-load
merchandise, warehouses, platforms, team tracks, and elevators
for receiving and forwarding car-load merchandise, and. perish-
able-freight hou.'^es and facilities for handling perishable freight.
Each railway as a rule has its own individual freight terminal
facilities, which are jealously guarded against use by a com-
petitor. This results in large areas in our commercial centers
being occupied, often extravagantly u.~ed, and susceptible to high-
er development in their use if shared jointly. A community of
interest in niany cases would prove beneficial to all concerned.
The possibility of future expansion of existing terminal facilities
will be a serious question in a few years and to avoid such a
contingency the question should be comprehensively considered
with a view to a modification in extent and use of the facilities.
The commerce of the city would be better served with a number
of joint freight terminals properly located in certain centers than
with large individual facilities remotely located and requiring long
trucking distances.
This question is a large one and surrounded with many
difficulties such as property rights, advantages of location, and
competitive conditions, which require the highest technical skill
to negotiate, — but it is of equal importance to the railways and
the commerce of the city, and should receive attention.
The matter of through freight traffic which afifects more
largely the railways has a bearing upon the community. This
traffic consists of such as that which passes through Chicago,
originating beyond and destined beyond Chicago. Only inasmuch
as it comes in conflict with and crowds out the facilities used for
local traffic does it affect Chicago. The handling of this through
trat^c is susceptible to great improvement, as has been previously
indicated by the writer. Much of it is now handled through the
congested centers of Chicago, to the detriment of local traffic,
and results in much annoyance from handling such traffic through
congested city districts. This traffic should be handled around
the city by means of belt railways with adequate interchange
yards, thus expediting the movement, reducing the cost of hand-
Ihig, and leaving the facilities in the inner districts for the hand-
Hng of local Chicago traffic. There is no more imperative need
in Chicago today than such an adequate interchange system for
through traffic exchanged among the thirty roads centering here.
It directly affects the revenue of every road and the welfare of
the community as well. A single remark made to the writer
recently by a large shipper in the northwest, that he could not ship
148 TH'E ARMOUR ENGINEER [Vol 4, No. 1
via the Chicago gateway owing to the delays and uncertainty of
movement through the Chicago district, should be a warning to
Chicago railways that this embargo should be removed, otherwise
the revenue would be seriously affected.
There is no railway terminal problem in this country as vast
and complex as that which exists in the Chicago district, but if
properly attacked and considered along common sense and scien-
tific lines by men trained and equipped for the purpose, it can
be solved with beneficial results to every railroad and the com-
munity as a whole.
THE ARMOUR ENGINEER
The Semi-Annual Technical Publication of the Student Body of
ARMOUR INSTITUTE OF TECHNOLOGY.
VOL. IV CHICAGO, JANUARY, 1912 NO. 1
Publishing Staff for the year 1912:
L. H. Roller, Editor.
M. A. Peiser, Business Manager. C. R. Leibrandt, Asst. Bus. Mgr.
Board of Associate Editors:
H. M. Raymond, Dean of the Engineering Studies.
L. C. MoNiN, Dean of the Cultural Studies.
E. H. Freeman, Professor of Electrical Engineering.
G. F. Gebhardt, Professor of Mechanical Engineering.
H. McCoKMACK, Professor of Chemical Engineering.
A. E. Phillips, Professor of Civil Engineering.
W. F. Shattuck, Professor of Architecture.
F. Taylor, Professor of Fire Projection Engineering.
Published twice each year, in January and in May.
Publication office: Thirty-third St. and Armour Ave., Chicago, 111.
TERMS OF SUBSCRIPTION
The Armour Engineer, two issues, postage prepaid $i oo per annum
Of this issue of The Armour Engineer IOOO copies are printed.
The technical press is invited to reproduce articles, or portions of
same, provided proper credit is given.
With this issue THE ARMOUR ENGINEER makes the
first appearance of its fourth year. The progress Avhich has thus
far been made with it in the short time of its existence is remark-
able, for in three years it has risen to a place
The Armour with technical college publications which have
Engineer existed for over twenty years. The credit for
and Its this is due in the greatest part to those alumni
Contributors who have given their support bv the contribu-
tion of articles ; the excellence and splendid
character of the subject matter of these articles have enabled
THE ARMOUR ENGINEER to take a front seat in the row
of college publications.
150 THE ARMOUR ENGINEER |Vol. 4, No. 1
That this standard of excellence can be maintained there is
no doubt, and in looking over the articles in this issue, we feel
sure that we have taken another step forward. We wish at this
point to thank those members of the faculty and the alumni who
have helped in making this issue what it is ; the preparation of
their articles has rec|uire(l the devotion of both time and money,
and in most cases the data contained in them has taken years to
collect, and has required vast experience and cost thousands of
dollars to produce.
From time to time THE ARMOUR ENGINEER has pub-
lished original articles on subjects which have lieen of great com-
mercial importance, and these articles have been reprinted and
abstracted in various publications in this country and Abroad, such
as Scientific American, The India Rubber World, Engineering
Record, India Rubber Journal of \jmd()n,Cumnii-Zeitung oi Ber-
lin. 71ie Chemical Engineer. Journal of Industrial and Engineer-
ing Chemistry of the American Chemical Society, etc. The fact
that the>e articles have been thus reproduced not only emphasizes
their excellence, but also shows the loyalty of their authors in
giving THE ARMOUR ENGINEER preference over such mag-
azines as those mentioned.
Anv)ther point which shows appreciation of the efforts of
our contributors is the fact that from time to time we have re-
quests for hack numbers containing certain articles. As a rule
we find that we cannot supply these copies because in the first two
vears uo provision was made, such as we are now making, for
keeping reserve numbers on hand to supply future demand.
We have received a communication from The Aero Club of
Illinois, the subject matter of which will prove of interest to many
of our readers, especially those who have more than a curious
interest in aeronautics.
The Aero Club It is the sentiment of The Aero Club of
of Illinois Illinois to keep in touch with the several aero-
nautical and engineering societies in this coun-
try, and with all individuals in the vicinity of Chicago who are
interested in aeronautics, such as there are at the Armour Insti-
tute of Technology.
Jan., 1912] EDITORIAiLS 151
Tiie Club is rapidly developing its various resources and
planiung some great things for 1912. Its Flying Field at Cicero,
just at the end of the Douglas Park branch of the Metropoli-
tan Elevated, is considered as one of the best in the world : it
now contains about a dozen machines, many of which are espe-
cially interesting because they incorporate principles of design
not previously applied in practice, but whose possibilities are
evident to the engineer.
Fortnightly sessions are held at the headquarters of the Club
in the Auditorium Motel, at which valuable technical discussions
take place, peculiar phenomena are illustrated, and matters of
special aeronautical interest are featured. Banquets are held
from time to time.
An invitation is extended to all Armour engineers to get
acquainted at the Club headquarters in the Auditorium, to visit
the aviation field at Cicero, and to make use of its facilities as far
as may be desired.
Prof. M. F>. Wells, associate professor of bridge and struc-
tural engineering at A. I. T., and Sydney V. James, '07, are
members of the Aero Club, an.d actively engaged in its work.
Anyone who is interested and desires to get further informa-
tion regarding membership, dues, etc., in The Aero Club, should
communicate with H. W. Robbins, Room 130, The Auditorium,
Chicago.
With the increasing material development in our civilization,
there comes an increasing need for knowledge relating to matter
and energy and to the methods by which these are utilized. This
need is felt not only by men engaged in en-
Engineering' gineering work but also by those working along
Education lines not usually considered to be closely re-
An Aid In lated to engineering.
Business The manager of a department store, for
Careers example, may be called upon to decide whether
to buy or generate the electric energy used in
the store ; to pass on a system of illumination ; to make a choice
of a iilan of heating and ventilating. Of course, in all of these
152 THE ARMOUR ENGINEER [Vol. 4, No. 1
cases, as well as in others of a similar nature, he would have
the recommendations of engineers ; and while he might follow
them fully, engineering knowledge would aid him in under-
standing their arguments for or against certain schemes and in
checking their final conclusions.
The business of banking is not fundamentally engineering,
but bankers supply money for engineering projects — the con-
struction of railroads, the development of mines, and the build-
ing of power plants. This money is not supplied, if the project
is a new one, without the advice of engineers, but those furnish-
ing the money can feel safer in making the investments if they
possess sufficient engineering information to reach independent
conclusions agreeing with their advisors.
Alany other illustrations might be mentioned to show the ad-
vantages of engineering knowledge in work that is fundamentally
not engineering, as that term is usually defined. However, it is
in business positions within companies whose work is largely of
an engineering character that the advantages of engineering
knowledge and ability show most abundantly. Salesmen for
products used in engineering work, superintendents of electric
railways or power plants, managers of manufacturing establish-
ments, all are very greatly benefited by engineering knowledge.
And it is being recognized in making promotions in such estab-
lishments that the engineer is a good man to put in a business
positon. The old route to the presidency by way of the office
boy is being abandoned and the engineer is being called upon
more and more to fill important business positions that were for-
merly occupied by men without engineering training.
E. H. Freeman.
At the December banquet of the Alumni Association of Ar-
mour Institute of Technology, Mr. Samuel Insull, President of
the Commonwealth Edison Company, delivered an address on
"Production and Distribution of Energy."
Mr. Insull 's Mr. Insull sketched briefly the develop-
Address on ment of central station business from its be-
Energy. ginning with generators of less than 100 kilo-
watts, to those now in use of 20.000 kilowatts.
Jan., 1912] EDITORIALS 153
The production of electric energy in large quantities has reduced
its cost to consumers ; and. in Chicaigo, a saving in copper for
distribution amounting to more than a million dollars could be ef-
fected if all the separate systems of distribution were combined in
one large system.
In addition to the economies which come from wholesale pro-
duction on a large scale, a further saving results from the fact
that different users of energy require it at different times. Less
machinery is necessary, therefore, when they are supplied from
one system.
The electrical operation of the terminals of the steam rail-
roads entering Chicago can be economically accomplished by cen-
tral station energy. The necessary power for the terminals of
one line having a large suburban traffic could be supplied by a sin-
gle generator at the Fisk Street Station.
Central station production of energy deserves special consid-
eration also on account of its lesseninjg the amount of smoke and
producing about twice as much electrical energy from a pound of
coal as is produced in small isolated power plants.
B. H. F.
The mid-winter dinner of the Armour Institute of Techno-
logy Alumni Association was held at the Hotel Sherman on the
19th of December, 1911, with eighty-five members present. In
the absence of Air. D. AlacKenzie, the master
The Alumni of ceremonies. Air. R. H. Rice acted in that
Association capacity.
The first speaktir was our president. Dr.
F. W. Gunsaulus, who gave a humorous and interesting address.
Mr. h. C. Fritch, Chief Engineer of the Chicago Great Western
Railroad, read a paper on "The Chicago Railroad Terminal Prob-
lem," which is given in full elsewhere in this issue. Air. Samuel
Insull, President of the Commonwealth Edison Company, de-
livered an address on "The Production and Distribution of En-
ergy," which is abstracted on another page.
After a short recess \'ice-President Harris called the meet-
ing to order, in the absence of President deBeers, and made a few
154 THE ARMOUR ENGINEER [Vol. 4, No. 1
announcements of interest to the members. The meeting was
then adjourned. ,
The (hities of the officers of any organization such as the
Alumni Association are at times extremely irksome and depress-
ing. Especially is this so of those of the corresponding secretary.
Several weeks before the alumni banquet
The of last month over 750 letters, with stamped
Corresponding return envelopes enclosed, were sent out to the
Secretary alumni of A. I. T. About 350 answers were
received, and thirty-five letters were returned
unopened. The supposition is that the remainder have l)een de-
livered and have received no attention.
The names of those whose addresses have been lost are given
below, and anyone knowing the location of any of these men will
confer a favor on the Alumni Association by supplying the ad-
dress.
GralT. H. W. (E) "00
Kaempfer, Albert (E) '03
Ouien, E. L. ( Ch ) '03
Raw son. Bovd H. (E) '03
Stillson. H. G. (E)'03
Coy, Frank A. (C) '04
Knapp, Morris T- (E) '04
^lac^Tillan. A. W. (AT) '05
RatcliiT. W. A. (M)'05
Snowden, C. R. (E) '05
Brock, W. L. (M) '06
Cutler, C. W. rEr06
Johnson, Carroll T. fAV06
Kimball, R. W. (AT) '06
Morrison, Ralph D. CM) '06
Nicholson, A'ictor (Ch ) '06
Pierce. Francis T. (C) '06
Smith, Geo. W. (U) '06
Boehmer, A. H. (M) '07
Nelson, C. J. fC) '07
Thompson, Morris fC) '07
Holmboe, Ralph (CV08
ATever, G. T. (C) '08
Pacvna. Arnold (Ch) '08
Jan.,
1912]
EDITORIALS
Pollak, Ernest
(C) '08
Thomson, F. L.
(FP) '08
Bexten, L. N.
(E) -09
Evans, R. T.
(M) '09
Perrine, A. A.
VE)'09
Schlinz, H. W.
(C)'09
Urson, Frank J.
(C) '09
Clarkson, Wm.. Jr.
(C) '10
Williams. D.
(C)'IO
Zeisler, Louis T.
(E) '10
Szeszvchi, T.
(C) "11
It may be interesting- to know that the Class of 1909 and the
present Senior class have appointed class secretaries. The idea
is to relieve the corresponding secretary of the Alumni Associa-
tion of some of the work by distributing it among the class secre-
taric^s.
156 THE ARMOUR ENGINEER [Vol. 4, No. 1
ARMOUR INSTITUTE OF TECHNOLOGY BRANCH OF THE
AMERICAN INSTITUTE OF ELECTRICAL ENGINEERS.
The Armour Institute of Technology Branch of the Ameri-
can Institute of Electrical Engineers exists to give the electrical
student an opportunity to hear and meet men prominent in elec-
trical engineering work, to make him familiar in discussing tech-
nical matters, such as those appearing in the Proceedings of the
A. I. E. E., and to enable him to talk and think on his feet, either
in presenting a paper or in discussing one.
The first meeting for the college year 1911-1912 was held on
September 21, 1911. The meeting was an informal one, as no
l)rogram had been prepared, and the chairman called upon several
members of the faculty and some of the students to address the
society. The responses brought forth a number of good im-
promptu speeches upon the purpose of the A. I. E. E., the value
of extemporaneous speaking, and the experiences of young men
in electrical work.
The second meeting was held on October 5, 1911. C. R.
Schuler, T2, read a paper on "Outdoor Arc Lighting." The
paper described in detail the series and multiple types of lamps,
and explained the different kinds of arcs, such as the plain car-
bon, the magnetite, and the flaming carbon arcs. The discussion
upon this paper was particularly interesting and instructive.
A special and joint meeting of the Branch was held with the
Mechanical Society on October 11, 1911. Dean Raymond pre-
sented a paper entitled '*A Few Hints to Engineering Students."
The Dean defined an engineer as one skilled in the application of
the forces and materials of nature to the uses of man. In his
paper he discussed the requisites of the young engineer, the engi-
neer and his work, the difficulties encountered by the young engi-
neer, and the present scope of engineering work.
The first November meeting was held on November 8, 1911.
H. P. Langstaff, T2, gave an illustrated lecture on "Power De-
velopment by the Winnipeg Electric Railway Co." The lecture
was a detailed description of the company's plant and distribution
system from the main hydro-electric generating station to the
smallest sub-station.
A joint meeting of the Branch was held with the Mechanical
and Civil Societies on November 21, 1911. Professor Wilcox
gave a lecture on "The Practical Application of the Gyroscope."
The lecture was a very clear and interesting description of the
fundamental principles and operation of the gyroscope. Many
cxami)les of the practical application of gyroscopic action were
mentioned and illustrated experimentally.
Jan., 1912] EXGIXEERIXG SOCIETIES 157
The first December meeting was held on December 13, I'Ml.
P. A. Strong, "12, presented a paper on "The Chicago and North-
western Railroad Signal Equipment." He described the operation
of the block signaling system as used by the Chicago and North-
western, and explained in detail the operation of the switch ma-
chines and signals as in use at present in the new Chicago and
Northwestern Station.
At the first January meeting, held on January 10. 1912. C. E.
Freeman, '97. Consulting Engineer, addressed the society on "Fi-
nancial and Engineering Features of Hydro-electric Develop-
ment." The address was very interesting and instructive, deal-
ing with the originating, financing, and engineering report of a
proposed development, and describing the work of the eno-ineer
in drawing up plans and specifications for the proposed develop-
ment when it is to become a reality. The matter of engineering-
supervision of the construction work was also discussed.
F . A. Graham.
ARMOUR BRANCH OF THE AMERICAN SOCIETY OF
MECHANICAL ENGINEERS.
The Armour Institute Student Branch of the American Soci-
ety of Mechanical Engineers has been decidedly successful thus
far this year, and the outlook for the future is very bright. The
Junior Mechanical class is exceptionally large, and its members
have already taken an active part in the afifairs of the Society. The
active membership of the Society is about thirty-five, and the
attendance at the meetings is increased by the facultv members,
who are always well represented.
At the Society smoker in October, Professor Gebhardt gave
an interesting talk, bringing out the purpose of the Societv ' and
illustrating its value as a training to the student. The purpose
of the Society is to train its members to speak before an audience
upon engineering subjects, rather than to present a course of
lectures. The intention is, however, to have occasional lectures
by men in actual engineering practice.
On N'ovember 1. 1911. Professor Frith gave an address on
"The Diesel Engine," explaining its construction and operation
with sketches and diagrams. Many good practical points were
brought out.
At the meeting on December 6, 1911, L. H. Philleo, '13, and
J. D. Bradford, '13, gave an interesting talk on a rotary gas en-
gine designed by themselves. Numerous drawings of the details
and of the assembled machine were exhibited, and a lively discus-
sion followed the talk.
158 THE ARMOUR ENGIXEER | Vol. 4, No. 1
The Society held its semi-annual banquet at Micheli's on De-
cember 20. V)l\. The attendance, including the Faculty of the
Mechanical Department, was fifty persons, and a very pleasant
evening was spent.
At the meeting held on January 11, 1912, Mr. A. W. Sem-
erak. T3, gave a talk, illustrated with lantern slides, on "Steel
Belting." This form of drive has not been developed in this coun-
try as yet, but in Germany steel belts are being manufactured and
used to a greater extent each year. They are giving good results
for both light and heavy drives. One of the marked advantage^
of steel belting is the reduced space requirement compared to
leather belts or rope drives. The steel belting used on the dirigi-
ble balloons of Count Zeppelin was quoted as a novel example of
an installation profiting by this feature.
P. L. Kcachic.
THE ARMOUR CIVIL ENGINEERING SOCIETY.
Although the Armour Civil Engineering Societ}- has had
some excellent speakers this year, the attendance at meetings ha>
decreased. It is hoped that during the coming year the members
will make an extra efl:'ort to attend the meetings, and that the
alumni and faculty will come out as often as possible.
At the first regular meeting of the college year, wdiich was
held in the engineering rooms on Tuesday evening, October 3,
1911, Dean Raymond addressed the society on "Helpful Hints to
Young Engineers." ^lany points were brought up which it is
well for a young engineer to knovv and bear in mind. The talk.
being appropriate and well rendered, was appreciated b}- all iires-
ent.
In order that the Seniors, Juniors, and the Faculty of the
Civil Department might have an opportunity of meeting one an-
other in a social way, the Society held a smoker on Friday even-
ing, October 13, 1911, in Chapin Club Rooms. The smoker was
well attended, and everybody joined in the spirit in whicli it was
given and had an enjoyable evening.
On Tuesday evening, October 17, 1911, W'm. Artingstall, of
the County Traction Company, addressed the Society on the
"Chicago River Tunnels." He illustrated with slides and working
drawings the completed work and that under way in the \^an
Buren, Washington and La Salle Street tunnels. Inasmuch as
Mr. Artingstall had worked on these tunnels himself, he was able
to deliver a very interesting and instructive talk.
At a joint meeting of the engineering societies of the college
in the Physics Lecture Room on the evening of November 21,
Jan., 1912] EXGlXEERiXG SOCIETIES 159
l*-*!!. Professor Wilcox delivered an instructive address on tlic
"Commercial Applications of the Gyroscope." By a series of
experiments Professor Wilcox successfully proved his theoretical
statements and illustrated many practical uses of the gyroscope
such as those in monorail cars, aeroplanes, and as an aid in in-
creasing the stability of ships.
The next meeting was held on Tuesday evening, December
5. 1911, when Stanley Dean. '05, addressed the Society on the
"Construction of the Hydro-electric Plant on the Hoosic River
at Schaghticoke, X. Y." Mr. Dean illustrated his talk with some
excellent slides, and as he had been "on the job" himself he gave
a very clear and concise talk. Each member was well re])aid for
attending the meeting, as it was one of the best of the vear.
On Tuesday evening, December 19, 1911, R. I. Randnlph
addressed the Society on the "Investigations of the Rivers and
Lakes Commission of the State of Illinois." Mr. Randolph told
of the number of men composing the commission, the number em-
ployed by the commission, its duties, and the work already accom--
plished, that under way, and that still to be done. As the lecture
was on an extremely interesting subject, and as it was well ren-
dered, it was well worth hearing.
C. IV. CoHiiis.
THE SENIOR CHEMICAL SOCIETY.
The members of the senior class in chemical engineering met
on September 18. 1911. and effected an organization by electing
officers to serve for the following year. It was decided at this
meeting that the society objects, the promotion of friendship
among the upper-classmen and the faculty, and the making of
acquaintances ^mong the older men of the profession, could not
be better carried out than by having monthly dinners, at whicii
some prominent chemical engineer or business man would be
asked to give an address on some topic of general interest, witli
which he was well accjuainted.
At the first of these dinners, held Wednesday, Octoi:)er 4,
1911, at the Union Restaurant. Prof. H. "McCormack, head of the
department of Chemical Engineering, was the guest of honor.
His address had for a subject "The Ethics of Chemical Engineer-
ing." The part that was specially emphasized as being the basis
of the talk was th^ statement, "Never give anyone advice upon
which you would hesitate to risk your own money." At the close
of the talk an informal discussion was held, during which most
of the faculty spoke. Mr. Pulsifer, the latest addition to the fac-
ulty, gave a highly instructive talk on mining conditions in the
160 THE ARMOUR ENGINEER [Vol. 4, No. 1
West. The entire faculty of the department and twenty of the
upper-classmen were present.
The second of the monthly dinners was held Thursday, No-
vember 2, 1911, at Kuntz-Remmler's restaurant. Mr. Hay, the
chief chemist of the Starck Rolling Mills Company, of Canton,
Ohio, was the speaker of the evening. He gave an excellent talk
on "Rust Prevention." He took up in considerable detail the
methods by which the sheet metal is treated in order to make it
rust resistant. The keynote of the process is good raw material
and careful treatment. The meeting was closed after a very ui-
formal discussion. The entire faculty and twenty-seven of the
students in the department were present.
The third of the monthly dinners was held Wednesday, De-
cember 6, 1911, at Kuntz-Remmler's restaurant. Mr. James S.
Sheafe, M. E., (Boston Tech.), Engineer of Tests of the Illinois
Central Railroad, delivered the address of the meeting, having for
a subject "The Function of Chemical Engineering in the Testing
of Railway Supplies." His talk was very good, ranging from a
description of the so-called reclamation service of the Illinois
Central Railroad, by which iron is re-rolled, brake beams re-
]:)aired, paint residues utilized, hose repaired, journal bearings
relined, etc., to a few words on the value of tact to an engineer.
The discussion which followed the talk was extremely interesting,
bringing out some very surprising facts as to the value of small
patent>. The attendance at this meeting was about the same as
at the other two.
Just at present the society is in the midst of preparations for
the semi-annual alumni banquet. This will be held Wednesday,
January 17, 1912, at the Sherman House. A very good program
has been arranged, the speakers being prominent alumni of the
society. Musical numbers will be given by the Chemical mem-
bers of the Armour Glee and Mandolin Clubs.
S. Kali II.
THE
ARMOUR
ENGINEER
THE SEMI-ANNUAL TECHNICAL PUBLICATION
OF THE STUDENT BODY OF
ARMOUR INSTITUTE OF TECHNOLOGY
CHICAGO, ILLINOIS
VOLUME IV. NUMBER 2
MAY, 1912
Copyright, 1912
BY
M. A. PEISER
THE ARMOUR ENGINEER
VOLUME IV. NUMBER 2.
MAY, 1912.
ROASTING AND SINTERING LEAD ORES.
BY H. B. PULSIFER.*
In the development of the metallurgy of lead to its present
standing in the modern industrial world three chief methods
have played a part in the art of reducing lead ores to base bul-
lion; we designate these as smelting with the ore hearth, smelt-
ing in the reverberatory furnace and reduction in the blast fur-
nace. Of these methods the one employing the reverberatory
furnace is all but extinct; the method of the ore hearth still per-
sists in a few localities for a particular ore, the smelting of the
silverless galena concentrates of Missouri, for example; but the
use of the blast furnace has constantly become of more and more
importance while at the same time undergoing remarkable im-
provement, increase in size and general efficiency until it is to-
day the standard and accepted method for the winning of lead
from its ores. It is of even more importance than this for with
the lead we are able to recover most of the silver, gold and cop-
per in the lead ore or in ores of these respective metals when
such ores can be used in connection with or to flux lead ores.
Now the conditions of reduction in the lead blast furnace
demand only enough sulphur left in the charge to form a matte
with the copper and some iron which is allowed for in making
up the charge; proper reduction also requires a strong reducing
atmosphere in the shaft of the furnace with a quiet and uniform
settling of the charge. It thus comes about that our blast fur-
nace charge can contain only a limited amount of sulphur, let
us say from three to five per cent ; any excess over this calcu-
lated amount must be eliminated before the material is sent to
the furnace. It is true that we have a constantly increasing pro-
portion of sulphide ores in our supply of lead ores, and, more
than this, these sulphides are usually concentrates. Fine ores are
ill adapted to treatment in a blast furnace; there is thus pre-
sented the double demand on preliminary treatment of an ore
intended for smelting, namely, to roast and to agglomerate into a
sinter.
It is to be noticed and clearly distinguished that the treatment
of lead ores has separated widely from the manner of treating
*Instructor in Metallurgy, Armour Institute of Technologry.
164 THE ARMOUR ENGINEER [Vol. 4. No. 2
copper ores. Roasting, and reverberatory furnaces and blast
furnaces are matters of high perfection in the metaUurgy of cop-
per, their deportment is httle related to the use of the same in-
struments when employed with a lead ore. The McDougall fur-
nace, which so admirably roasts fine copper concentrates for the
reverberatory, is not used on lead ores ; that grim and monstrous
affair with its lake of molten rock, which presents itself as the
copper reverberatory of to-day, has no meaning in connection
with lead ; likewise, the great blast furnaces devouring rock and
ore and coke into the stream of rainbow flames are utterly re-
moved from dark, smoothly sighing furnaces which are best for
lead ores. The best conditions for the treatment of each metal
have become widely differentiated, especially during more recent
years.
Thus it is that roasting lead ores has become an important
matter only since lead blast furnaces have become something
more than the mere adjunct to hearth and reverberatory smelt-
ing, which they once were. With the gradual development of
lead blast furnaces the long hearth reverberatory was for many
years the standard for roasting both lead ores and matte. Fur-
naces of other types came into limited use, of these the Brueck-
ner cylinder had a flashy career ; then the work of Hjmtington
and Heberlein gained recognition and started a new development,
revolutionizing the treatment of lead ores and directly leading to
that process which is at this very moment being budded into the
metallurgy of iron — the sintering of iron ores and iron flue dust
— a beautifully simple process which may well quiet all alarmist
conservationists by multiplying our available iron ore reserves
many times.
The long hearth reverberatory roasting furnace to a con-
siderable degree embodies the elements necessary for the success-
ful roasting of lead ores. In this type of furnace the air supply
can be sufflciently abundant, there is opportunity for an unlimited
amount of stirring, while the heat may be regulated as desired, a
gentle heat for the highest lead ores or possibly more for ores
less fusible. The furnace is not unreasonable either in first cost
or maintenance ; it is, however, costly both as to fuel requirement
and labor, while its capacity is not large.
Exact details will of course be different at each locality but
the ordinary furnace from forty to sixty feet long and about
fourteen feet wide, hearth area, will roast some twelve tons per
twenty-four hours, twenty-five per cent of the weight of the
charge, if coal is used for fuel, will be required ; the labor cost
can hardly be less than one dollar a ton. A battery of six of
these roasters might require labor as follows :
May, 1912] PULSIFER: LEAD ORES 165
Each shift:
1 Hoistman, @ $2.35 .^ 2.35
1 Trammer, @ $2.10 2.10
12 Sidemen, @ $1.90 22.80
2 Firemen, (a^ $2.10 4.20
1 Foreman, @ $3.00, 1-3 services 1.00
Total $32.45
If three shifts per twenty-four hours are used and each fur-
nace roasts twelve tons of material the cost is evidently
32.45 X 3
= 1.3o
12X6
or $1.35 per ton.
The roasting of the matte produced in lead smelting re-
quires the same careful heat and constant stirring that a lead ore
would. Mattes are thus customarily roasted in this same kind
of furnace ; in all cases the material is put in the furnace at the
flue end and gradually worked forward to the hotter region by
the labor of the side men. Excessive heat will fuse the charge,
metallic lead may form and with the fused sulphides gradually
creep into the hearth, raising the floor as the mass accumulates.
One way to remove these masses is to clear the floor, raise the
heat of the furnace for a few days and then tap from the side
of the furnace after driving a bar through the wall. Huge sows
of matte and galena are awkward afl^airs to handle with the
equipment usually available in a roaster shed. Sledging has little
effect upon them and drilling for floating is tedious. ^lelting,
either before or afer removing the furnace floor, commends itself
as the quickest way of getting them out. I knew of one
which kept growing and finally raised the floor of the fur-
nace so close to the roof that "the output of the furnace
was greatly diminished. During the digging out of the furnace
the output was of course large enough, but unfortunately, the
material was' not well roasted.
The design of furnace as built by the Colorado Iron Works
shows the long hearth and the fuse box at the grate end. Many
furnaces for the roasting of ores and matte, according to present
American practice, have the simple straight hearth, only.
This agglomeration or even fusing at the end of the roast-
ing, after the sulphur is largely removed, is to furnish a lumpy
product for the smelting in the blast furnace which is the next
step in the ore treatment. The use of the fuse box requires
added labor and entails increased loss of metals by volatilization.
166
THE ARMOUR ENGINEER [Vol. 4, No. 2
May, 1912] PULSIFER: LEAD ORES 167
This idea of sintering the materials has long been standard prac-
tice. Percy mentions various places on the Continent where the
lead ores were thus treated; at Freiberg the "Sinter-calciners"
had part of the hearth on a higher level than the portion where
the final treatment took place. "Thus each charge undergoes cal-
cination during sixteen hours. The temperature should be kept
high enough to cause the charges near the fire-bridge to sinter
strongly and become pasty^ in which state they are fitted for the
smelting process;" at Pontgibaud the furnaces w^ere double
decked with fuse box ; Mechernich had a battery of ten furnaces,
each with its hearth thirty-two by twelve inches.
All the text-books on the subject give numerous and inter-
esting details in connection with this kind of furnaces; an ap-
proved type is the furnace built by Eraser and Chalmers.
There is a goodly volume of literature available about re-
verberatory roasters for lead ores; we can hardly consider in
detail the matters of draft, materials of construction, design of
parts, methods or schedules of rabbling and drawing charges, the
complicated reactions during roasting, the sulphur elimination or
the practice at numerous plants, etc.
With cheap labor, cheap fuel and twelve-hour service, Col-
lins states that the cost of roasting and fusing should not exceed
eight shillings per ton. In our western states these conditions
are seldom fulfilled and costs may more reasonably be placed
around three dollars a ton.
Since the development of blast-roasting any further prog-
ress in connection with reverberatories is evidently out of the
question. The matter of labor is always serious ; the work is ar-
duous and exhausting, it is a continual struggle to keep the stand-
ard high and sulphurs low in the product. The loss by volatiliza-
tion of both lead and silver values is always considerable and
precludes the richest ores being treated this way. Handling the
flue dust and fume is a serious matter while sulphur trioxide is
always formed and usually needs particular attention. By the
method disclosed in the Sprague patents the gases can be neu-
tralized in their sulphuric acid content and all fume recovered ;
the successful operation of this method indicates quite unusual
metallurgical attainment.
Possibly no better resume of the status of reverberatory
roasting has been given than that by Arthur S. Dwight in his
recent paper on the "Efficiency of Ore Roasting." The paper
appeared in School of Mines Quarterly for November, 1911,
and was shortly afterwards reprinted in the Engineering and
Mining Journal and abstracted in Mining and Engineering World.
168
THE ARMOUR ENGINEER
[Vol. 4, No. 2
May, 1912] PULSIFER: LEAD ORES 169
Extensive installation for the treatment of ores by the use
of the Broeckner was made at Murray, Utah. A descriptive ac-
count is given in Engineering and Mining Journal, 1907, page
527 and 575, by W. R. Ingalls; it is stated that there were twenty
cylinders, each eight and one-half by twenty-two feet. Whatever
success (see Collins. Metallurgy of Lead, page 102) may be
credited to these roasters the introduction of blast-roasting has
put them practically out of date, — none of those at Murray were
used during 1909 or 1910. The illustration is of a furnace once
extensively advertised by the Colorado Iron Works.
Brueckner cylinders labored under the disadvantage of mak-
ing much flue dust and being difficult to keep in condition. The
heat also required very careful regulation.
The heap roasting and stall roasting of lead ores has been
of small importance, In the metallurgy of copper it has been
of far greater moment ; the same may be said of mechanically
raked reverberatory furnaces and the various types of revolving
and multiple hearth furnaces. The character of ore and require-
ments to be met are altogether different and have brought splen-
did results in this other field.
It is a principle of the most vital importance that for oxi-
dizing effects by the use of air the material should allow the air
to pass through and not simply oirr its substance ; constant stir-
ring of the material in contact with air is a poor substitute, it
is, however, far better than to allow the material to lie inactive.
The present mechanical furnaces for roasting copper and zinc
ores are based on the principle of constant stirring; yet, if any
means should be devised to operate as perfectly mechanically and
at the same time submit the ore to a current of air passing
through, instead of over, the mass, the present lurnaces would
quickly be superseded.
The present success of blast-roasting, or "sinter-roasting"
to bring Percy fully to date, began with the work of Huntington
and Heberlein in Europe, which men first blew air through the
material (lead ores) to be roasted ; Carmichael and Bradford al-
tered the composition of the charge, Savelsberg changed the
charge still more; Greenawalt devised the method of supporting
ores on an inert bed while drawing air down through by suction
and, finally, Dwight and Lloyd made the process continuous. W<
must remember that, "Sinter-calcining" is a term used by Percy
in 1870 and that the desirability of both roasting and sintering
lead ores was even at that time perfectly understood. Roasting
and agglomerating lead ores in the reverberatory furnace has al-
ways a desirable method of preparing these ores for the reduction
in the blast furnace, but this preliminary fusing together is al-
170
THE ARMOUR ENGINEER [Vol. 4, No. 2
p
^^Bi'j'l|%f^jHHMHiiHBH^S
N|
9
May, 1912] PULSIFER: LEAD ORES 171
ways more costly and with rich ores hardly to be attempted on
account of lead and silver losses ; according to Hofman ores con-
taining more than twenty per cent lead should not be heated
hot enough to fully agglomerate them. It may also well be
kept in mind that neither up-blast nor down-suction treatment
his in any instance been the special point of improvement or in-
vention claimed for any of the processes about to be described.
Both are old practices in the treatment of ores.
Huntington and Heberlein found that by mixing lime with
rich galena ore and first partially roasting in a mechanical fur-
nace this intermediate product could be both suitably roasted and
agglomerated by blowing the still hot mass in a convenient re-
ceptacle ; the original claim of their United States patent is thus :
(U. S. Patent 600,347, March 8, 1898.)
"1. The herein described method of oxidizing sulfid ores
of lead preparatory to reduction to metal, which consists in mix-
ing with the ore to be treated an oxid of an alkaline-earth metal,
such as calcium oxid, subjecting the mixture to heat in the pres-
ence of air, then reducing the temperature and finally passing
air through the mass to complete the oxidation of the lead, sub-
stantially as and for the purpose set forth.
"2. The herein-described method of oxidizing sulfid ores of
lead preparatory to reduction to metal, which consists in mixing
calcium oxid or other oxid of an alkaline-earth metal with the
ore to be treated, subjecting the mixture in the presence of air
to a bright red heat (about 700 degrees centigrade), then cool-
ing down the mixture to a dull red heat (about 500 degrees
centigrade), and finally forcing air through the mass until the
lead ore, reduced to an oxid, fuses, substantially as set forth.
"3. The herein described method of oxidizing lead sulfid in
the preparation of the same for reduction to metal, which con-
sists in subjecting the sulfid to a high temperature in the pres-
ence of an oxid of an alkahne-earth metal, such as calcium oxid,
and oxygen, and then lowering the temperature, substantially as
set forth."
Seven years later they took out another United States patent,
786,814, of April 11, 1905, after others had proved that good
results could be gained without the use of lime. They, of course,
retained their preliminary roasting and final blowing. The fol-
lowing quotation from the patent indicates their change of
mind regarding the use of lime as well as their recognition of
the desirabiHty of making a good sinter:
"To increase the fusibility of the ores treated, limestone, sil-
ica, or oxid of iron may be added ; but this is not in all cases
necessary. Indeed, one of the advantages resulting from sud-
172
THE ARMOUR ENGINEER [Vol. 4, No. 2
May, 1912] PULSIFER: LEAD ORES 173
denly and greatly reducing the temperature with the aid of water
is that the addition of calcium oxid or other suitable oxid, as
described in the specification of our aforesaid patent and as
mentioned above, may in many instances be dispensed with."
The fourth and final claim of their patent is as follows :
"4. The herein described process of oxidizing sulfid ores
preparatory to their treatment for the reduction of the metal
contained therein, such process consisting in heating and working
the ore until the proportion of sulfur contained therein is re-
duced to twelve per cent or thereabout, rapidly cooling the ore
with water down to atmospheric temperature or thereabout, re-
starting combustion in the mass and forcing air there through
for the purpose of further desulfurizing and oxidizing the same,
substantially as set forth."
This work of Huntington and Heberlein has not only been
of enormous importance because of the magnitude of the work
carried out under these specific patents, but because of the in-
centive for the other developments which came directly in its
wake. Huntington and Heberlein were seriously mistaken about
the chemical reactions involved in the operation of roasting
and sintering, but gained eminent success in their method
of treating the material. They put the mechanical roasting
furnace to just that limited use it is susceptible of in roast-
ing lead ores — a light preliminary roasting, eliminating only
part of the sulphur and avoiding all sintering. Later
practice with other methods avoids preroasting by diluting the
original sulphides with materials not containing sulphur ; in
this way a charge can be prepared which shall contain the right
sulphur content, or fuel value, for successful roasting. Pre-
liminary roasting will remain a standard method for reducing
the caloric value of ores where diluting is not desirable. In
view of the fact that it not only wastes the inherent fuel value
of the ore, but requires additional extraneous fuel to accomp-
lish this, it is very uneconomical. It is desirable to here emphasize
the significance of process in general — that after a compara-
tively light treatment in a mechanical roasting furnace the ma-
terial is blown in a converter and in one operation both de-
sulphurized and agglomerated, or sintered. The success of the
process marked a great advance in metallurgical progress.
A fairly complete list of the papers descriptive of the
process will be found in an article by the author in Metallurgical
and Chemical Engineering for March, this year. Besides those
relating to the practice as carried out in the different smelteries,
quite a few of them relate to the chemical theory. The most
recent paper along this line is one by C. O. Bannister, presented
174
THE ARMOUR ENGINEER [Vol. 4, No. 2
May, 1912] PULSIFER: LEAD ORES 175
to the Institution of Mining and Metallurgy, February 15, 1912,
entitled, "On the Theory of Blast-Roasting of Galena." It
records some of the best laboratory experiments yet under-
taken and the conclusions are in line with our recent concep-
tions of the reactions involved.
The Carmichael-Bradford process, as once carried out at
Broken Hill, New South Wales, offers its chief interest in that
it was one step removed from the use of lime as a necessity in
the blast-roasting. Converters similar to those used in the
Huntington-Heberlein process were employed and the gas was
rich enough in sulphur dioxide for converting into sulphuric
acid. In dispensing with the preliminary roasting it ought to be
stated that the composition of the charge at Broken Hill was
far from that represented by a rich galena, much rather ap-
proaching that at present used in the latest practice where
sulphurless diluent is added. The following paragraphs are from
the United States patent No. 705,904, July 29, 1902.
"This invention relates to the treatment of sulfid ores or
metallurgical products preparatory to smelting, and more par-
ticularly to the treatment of lead such as sulfids of zinc, copper, or
iron and mixtures of the same. Its object is to desulfurize and
cinerate such ores or products and to change them into a condition
more suitable for the smelting process. In the processes hitherto
employed such ores or products have first been subjected to a
preliminary roast in a suitable furnace, with or without suit-
able chemical reagents, for the purpose of oxidizing or sul-
fating the metallic sulfids. The charge has then been removed
to a converter or other suitable receptacle, where it has been
subjected to a current of induced air for the completion of the
oxidation.
"My invention consists, essentially, in dispensing with the
preliminary roasting and in the complete desulfurization and
cineration of the sulfids in the converter in one operation. I
accomplish this by mixing with the raw sulfid ore or metal-
lurgical product a suitable proportion of calcium sulfate and by
subjecting the mixture in a converter to the action of an in-
duced current of air, starting the reactions by means of heat,
whereby sulfate of the metal and the calcium sulfid are produced,
and the calcium sulfid in its oxidation produces sufficient heat to
set up the necessary desulfurization reactions and to thoroughly
oxidize and cinerate the ore without loss by volatilization."
The Savelsberg process attained wider distribution than
the Carmichael-Bradford, which was limited to Australia, be-
cause it used limestone in place of gypsum as diluent. W. R.
Ingalls had an article in the Engineering and Mining Journal
176
THE ARMOUR ENGINEER
[Vol. 4, No. 2
on this process; 1905. Vol. LXXX, page 1067; also a short
editorial with three excellent illustrations in the same journal
in 1906, Vol. LXXXI, page 1136. Although used some years
in this country and Europe it has never become as important
as the Huntington-Heberlein process, being already partly sup-
planted by the Dwight-Lloyd machines. A couple of clippings
from Savelsberg's United States patent No. 755,598, March
22, 1904, are interesting. The charge is introduced in layers,
at intervals, into the pot as the roasting progresses and the final
Dia^rnin of Savelsberg Converter.
cake is discharged by tilting the pot. The success of the process
has been most pronounced when treating rich galenas not con-
taining pyrites.
"My invention is based on the observation which I have
made that if the lead ores to be desulfurized contain a sufficient
quantity of limestone it is possible by observing certain pre-
cautions, herinafter set forth, to entirely dispense with the pre-
vious roasting in a roasting furnace, hitherto necessary, and to
176
THE ARMOUR ENGINEER
[Vol. 4, No. 2
on this process; 1905, Vol. LXXX, page 1067; also a short
editorial with three excellent illustrations in the same journal
in 1906, Vol. LXXXI, page 1136. Although used some years
in this country and Europe it has never become as important
as the Huntington-Heberlein process, being already partly sup-
planted by the Dwight-Lloyd machines. A couple of clippings
from Savelsberg's United States patent No. 755,598, March
22, 1904, are interesting. The charge is introduced in layers,
at intervals, into the pot as the roasting progresses and the final
Diagram of Savelsberg Converter.
cake is discharged by tilting the pot. The success of the process
has been most pronounced when treating rich galenas not con-
taining pyrites.
"My invention is based on the observation which I have
made that if the lead ores to be desulfurized contain a sufficient
quantity of limestone it is possible by observing certain pre-
cautions, herinafter set forth, to entirely dispense with the pre-
vious roasting in a roasting furnace, hitherto necessary, and to
The Armour Engineer,
IV— 2. May, 1912.
Roasting and Sintering Lead Ores,
H. B. Pulsifer.
General Arransenient of Slnterine Plant at Sallda, Colo. .
May, 1912] PULSIFER: LEAD ORES ^ ' 179
desulfurize the ores in one operation by blowing air through
them.
"Liquefaction of the ores does not take place, for although
a slag is formed it is at once solidified by the blowing in of the
air, and the passages formed thereby in the hardening slag al-
lowing of the continued passage there through of the air. The
final product is a silicate consisting of lead oxid, lime, silicic
acid, and other constituents of the ore, which now contains but
little or no sulfur and constitutes a coherent solid mass which
when broken into pieces forms a material suitable to be
smelted."
The development of the Huntington-Heberlein, Carmichael-
Bradford and Savelsberg processes all meant more than simply
better and cheaper methods of roasting and sintering lead ores,
three other results stand out as strong steps in the advance of
metallurgical practice, namely, less volatilization, decreased
amounts of sulphur trioxide in the gas and using the sulphur
content of the ore as fuel without resorting to other and costly
source of heat.
It is not to be understood that any one of these steps was
complete or final but an advance in the right direction. Under
the intense and long continued heat in the hand-raked reverbera-
tory losses of lead and silver were too great to allow the sintering
of many commercial ores, they were roasted as well as possible
and charged into the blast furnace although much too finely
granular ; in the roast pots and under the influence of the forced
blast the reactions take place much more quickly and losses are
considerably reduced. If the pots are working well no one por-
tion of the charge is hot all of the time, the oxidation and sin-
tering progresses upward in more or less of a horizontal plane
with fairly cool sections above and below, this is a distinct aid
in keeping the amount of fume low.
In the same way with regard to the production of sulphur
trioxide, that substance which on contact with water forms sul-
phuric acid, the prolonged and repeated contact of hot ore and
hot gases formed sulphates in the reverberatory furnace, on
breaking down at the time of slagging the sulphur trioxide lib-
erated escaped through the flues and entered the air to devastate
vegetation, or, if caught in a bag house, of necessity neutralized
and rendered innocuous before coming in contact with the bags.
By roasting in pots the quick reaction and localized hot zone has
much decreased the amount of this oxide of sulphur which is
formed. It is true that with the removal of the sulphur an oxide
is formed and that this oxide escapes into the air ; but it is the
dioxide, a substance far dififerent from the trioxide, it has no
180 - THE ARMOUR ENGINEER [Vol. 4, No. 2
particular harmful effect upon woolen bags nor upon vegetation
when diluted with air to that extent which naturally comes when
delivered into the air from the top of a stack two hundred or
more feet high.
The matter of using the inherent fuel value of the ore is
great economy, practically it is a matter of many dollars and
cents, and in practice, too, it is a big step in the art of control
and in the art of guiding the inherent forces in materials. One
sighs in relief that we have at last partly conquered the refrac-
tory sulphides ; could the sulphides realize, they, too, would
likely be relieved that they are no longer outraged in rever-
beratories but allowed to express their desired reactions in pots
and under the influence of the gentle blast.
The roast made in converter pots is better than reverbera-
tory roast because it smelts better, — in particular it smelts faster.
It is much cheaper to produce because it requires less hand
labor and less extraneous fuel. At first the pots were rather
small, holding only a ton ; with proficiency the size was increased
until standard practice now uses converters__ holding eight tons,
while even larger have been built. If the charge is pre- roasted
fuel will be required for this part of the process but none in the
pots ; sometimes the bed of the charge is made of hot pre-roast
with a layer of diluted sulphides on top; sometimes the whole
charge is of diluted sulphides, in which case fuel is required in
the bottom of the converter to start the reaction. This latter is
the common practice at ]\Iidvale, Utah, where the converters
have the form of stalls instead of pots, when the charge is
roasted the front of the stall is raised and the sintered cake is
pushed out from behind with a ram.
The cost of labor is still the main item of expense; it is now
become labor to operate machinery instead of using labor as a
source of power. With the reverberatory furnace men toiled
with heavy rabbles by the side of the hot furnace to stir and
push the ore along the red hot hearth, a few pounds at a time;
with converters the ore is dropped in with chutes or from hop-
pers, men level ofT the charges and break up blowholes if they
develop but the cake is removed bodily by tipping or pushing or
a crane picking up the pot and all and dumping the cake out bod-
ily. The cake is broken by dropping or by dropping a weight
upon it until the pieces will enter the large crushers provided,
which crush to the size required for blast furnace work.
All the good things we have said of blast-roasting in pots or
in stalls are to be intensified in regard to sinter-roasting as may
be done with down-draft practice in continuous machines or in
pans. Down-draft means sucking air down through the ignited
May, 1912]
PULSIFER: LEAD ORES
181
charge by means of an exhaust fan. For this work the charge
is best in a fairly thin layer, not over ten inches thick, and as the
discharge may take place by merely freeing the cake at the end
of the continuous machine or by inversion of the pan it is no
reversion to manual labor.
The sintering of the charge is accomplished even more satis-
factorily than in converters, sulphur elimination is equally good,
volatilization is less, evolution of sulphur trioxide probably like-
wise less while total operating cost is decreased.
To Dwight and Lloyd we owe the development of the prin-
ciple, of the continuous operation while to Greenawalt is due
particular credit for showing us how to roast with down-draft
I
^^^5(9^5
^I^Miti^tJgt^gljif^i^i^l^^
Diaj^raiii of au Annular Dwight-LIoyd Continuous :\Iaohine.
on cast iron grates and for much research in connection with the
operation of intermittent pans.
In the Greenawalt patent No. 839,065, issued December 18,
1906, the use of suction acting downward in connection with a
porous bed is clearly specified in these words (claim 5) :
5. "An ore-treating process consisting in placing the ore
upon a porous bed, subjecting the same to heat, and passing the
resulting fumes or a portion thereof down through the bed by
suction acting from beneath the bed."
This patent was the result of work on trying to improve the
efficiency of zinc roasting furnaces, but was broad enough to
cover the case of treating other ores as well. It is apparently a
chronic disposition of patents to aspire to usefulness in fields not
182 THE ARMOUR ENGINEER [Vol. 4, No. 2
at the time of particular significance but which may be fertile a
little later ; it happened thus with this patent for improving zinc
roasting died but the development came with lead ores and we
find the Dwight-Lloyd patent No. 882,517 of March 17. 1908,
virtually an incorporation of the Greenawalt idea. Claim 12 in
this patent reads as follows :
12. "The process for roasting finely divided ores or metal-
lurgical products containing sulfur or other combustible elements
and sintering or agglomerating the particles of the roasted ma-
terial into a coherent mass through the action of the heat gener-
ated by internal combustion, which consists in disposing said
fine material in a layer upon a support, substantially as set forth,
whereby there are gas exit passages provided below the said
layer, igniting the material at its upper surface, causing a current
or currents of suitable oxidizing gas to pass downward through
the said surface and under uniform distribution over the same
and to pass thence in a downward direction through the layer
and through the gas exits below it, whereby the combustion is
carried from said upper ignited surface downward through the
mass to the lower surface thereof, maintaining all particles of
the mass under treatment in a relatively quiescent state, whereby
is effected the complete sintering together of the roasted par-
ticles into a coherent cake, and finally removing the cake or sin-
tered mass independently of its supporting or holding devices,
substantially as set forth."
A patent granted to Charles Vattier December 26. 1893 (No.
511,476), embodied many of the principles of Dwight-Lloyd
sintering, especially the restraining of exit surface, the other
ideas of sintering had been either long practiced ("sinter-cal-
cining" was a term used by Percy in 1870) or specifically em-
bodied in the patents of Huntington and Heberlein, the Car-
michael-Bradford patent, the Savelsberg patents, the patent of
Greenawalt and that of Herbert Haas (No. 808,361, IDecember
26, 1905). The really meritorious invention of Dwight and Lloyd
came to notice in their second patent. No. 882,518, March 17.
1908, which expounds the idea of the continuous machine.
Attempts to restrain a surface and force a current of air
through are apparently ill taken, it makes no diiiference whether
the air is sent up through from below or forced down from
above, channeling inevitably begins at the surface of entrance
and breaks out on the opposite surface ; restraining the surface of
exit can only partially allay the difficulty. Forcing the air down
through the charge and out through the grate was tried at Port
Pirie, Australia, where the results were disastrous, due to too
much channelling and irregular sintering of the charge. The
May, 1912] PULSIFER: LEAD ORES 183
fundamental difference is that a push or pressure in the yield-
ing charge opens a channel, suction draws the mass together
and closes the would-be opening.
The way by far best of all to get uniform and complete
smtenng is to ignite the upper layer of the charge and draw the
blast and combustive layer down through to the grate. This has
been embodied in the continuous machines and intermittent pans.
For the protection of the grate a porous layer of inert, incom-
bustible material such as limestone, iron ore or roasted sinster
is absolutely necessary. If such is not used the grate will be com-
pletely and quickly ruined by the corrosive action of the hot sul-
phides.
The operations involving the continuous charging, ignition
roasting and sintering and discharging as exemplified in the
Dwight-Lloyd machines are very well worked and reflect crreat
credit upon their inventors. The cut shows the arrangement of
the plant at Salida, Colorado; a full account of this plant is
given in Metallurgical and Chemical Engineering, February 191?
whence our illustration, by permission. (See insert, page 177.) "'
The continuous machine patents described three types cylin-
drical, annular and straight line; the cylindrical machine has had
Its photograph widely distributed in recent literature but is evi-
dently out of favor at present, the annular type is apparently the
favorite one in Europe while in this country the straic^ht line
type IS the one now generally installed. Serious points of oper-
ation are naturally the uniform charging or loading of the pallets
as they come beneath the charging hopper, the loss of suction by
leakage and the obstruction of pipes and fan with the sticky
sulphurous fume which is given off during the roastino- Pro-
vided the porous bed is of sufficiently inert, or better yet matte
absorbing material, and uniformly distributed on the grates be-
neath the charge proper there should be little difficulty in dis-
charging the finished cake. Conversely, the more the hot sul-
phides actually touch and fuse to the grates so much more is the
difficulty in getting the product removed.
The fan has been a particular source of difficulty with all
suction work. Not only must the fan handle a large volume of
gas efficiently, but with a considerable actual pull or vacuum
No type of fan commonly found in the market fills the re-
quired conditions of size, capacity and pull. The accompanying
sketch shows the outline of the special fans as built for the
Salida plant by the American Blower Company. It is neces-
sary to water cool the bearings while for the speed required
the center hung wheel is essential, it must also be of the strong-
est and most rigid construction. Large sections of the scroll
184
THE ARMOUR ENGINEER
[Vol. 4. No. 2
are so assembled that they shall be easily removable for cleaning
out the interior of the fan.
It is not yet recorded that the continuous machines have
been very successful in sintering ores as high in sulphur as
those commonly treated in converters or stalls ; thus while
eighteen per cent is given as nominal for the older practice
the mixture for continuous machines is run at about fourteen
per cent at Bindfeldshammer (Guillet, Revue de Metallurgie,
Diagram of American Blower Fan Used at Salida.
August, 1911), and at about fifteen per cent at SaUda (Metal-
lurgical and Chemical Engineering, February, 1912), in each
case the excess of sulphur is first removed by partially roasting
in a mechanical furnace.
The construction of a Dwight-Lloyd continuous machine is
essentially an endless grate which is continuously undergoing
the operations of being bedded, loaded, ignited, sintered and
discharged; each of which separate functions is progressing
May, 1912]
PULSIFER: LEAD ORES
185
without interruption at its own proper part of the machine while
the grate progresses slowly in its course, around and around.
The parts and functions have to be so proportioned that
the grate shall be properly protected against corrosion by matte
and slag, the layer of coarse or inert material next the iron is
for this purpose ; its necessity is greatest with the more easily
fusible charges and charges which are high in sulphur or other
fuel which will afford much heat and fusion of the charge. The
ignition is to be so regulated that the charge shall be properly
started toward self-sintering yet avoiding any excess, which
Diag-ram of Dwight-Uoyd Cylindrical Machine.
would be wasted. The suction and size of the suction box be-
neath the grates is likewise so adjusted that after the proper
course of the roasting and sintering is accomplished the cake
is promptly beyond the cut-off and advancing toward the dis-
charge. Of course the speed at which the grate travels affords
a means of regulating all these details to a considerable extent.
As details relating to the cost of operation, in particular
the cost of repairs and renewals, covering considerable periods
are not available much interest will attach to the performance
of the machines as reported from time to time.
186
THE ARMOUR ENGINEER
[Vol. 4, No. 2
The intermittent dumping pan type of unit we choose to
speak of in connection with the name of Greenawalt because
he has done much work not only in designing practical equip-
ment but because he has so thoroughly studied the composition
of charges and how to sinter successfully. It is not to be sup-
posed that he originated the dumping pan, but he was probably
the first to appreciate and use a bed of protecting material be-
o.=Qa
XlzZZZZ^VTZZZZX
Fan Designed by Greena^valt for Down-Draft Sintering.
tween the charge and grate. The matter of thorough mixing
of the charge and appropriate moisture content, points all too
often neglected by others, were fully understood and turned
to account by this metallurgist.
The dumping pan has a good deal of merit when compared
with continuous machines, its capacity is likely less per unit,
but original cost of a unit is likewise less while the mechanical
May, 1912] PULSIFER: LEAD ORES 187
operation is about as simple as can be imagined. This is of the
utmost importance in metalhirgical work. About ten inches has
been found a suitable thickness for the cake, this allows ade-
quate protection of the grate without making the proportion of
bed to charge unduly great. The charging, ignition and suction
requirements of the pan have also been pretty well worked out
and prove decidedly practical on a commercial scale.
The illustration of a fan as designed by Greenawalt gave
general satisfaction, the bearings, however, would be better made
even stronger with end thrust better provided for. The fan
should also be split horizontally about half way up so that the
casing can be raised on a hinge and the whole interior exposed
for cleaning. This is necessary if the fan is placed in close
proximity to the roasting unit for the gases from roasting are
never very hot, at the beginning of a heat they are heavily laden
with water vapor and will usually contain considerable amounts
of elemental sulphur, lead and arsenic fume. If this all has to
pass through the fan before depositing in a flue an amount quite
inappreciable will be sufficient to seriously interfere with the
capacity of the fan. With proper arrangements no longer than
ten or fifteen minutes a day interruption of the suction should
suffice to keep the fan clean and in order.
During any one heat in the pan the suction starts with air
passing quite freely and a vacuum which is much less than will
be produced as soon as the roasting is well under way with con-
densation of water in the lower part of the charge. Little by
little the passing of air diminishes as the charge evidently gets
filled with water in its lower part, maybe only a tenth part of
the starting amount will soon be passing ; but at the same time
the charge is getting hotter below the fire line and it is not very
long before the minimum amount of air is passed through, and,
with the drying out, the charge becomes more and more por-
ous, considerably before finishing the charge is even more porous
than at first and at the end from two to three or even five times
that amount may go through. With this increasing permeability
the intensity of the suction falls ofif equally and the gases reach
their maximum temperature.
The generation and absorption of the heat is an interesting
phase of the operation ; unfortunately I do not find data for
the heat of decomposition of pyrite nor for the heat of forma-
tion of lead silicate, in neither case are the quantities large and
we may hope they balance nearly equally. It will be seen that
about one-third of the heat generated is absorbed in the charge;
of the remainder a portion is radiated and conducted away
from the top surface of the charge during the first few minutes
188
THE ARMOUR ENGINEER [Vol. 4, No. 2
May, 1912] PULSI'FER: LEAD ORES 189
of the run, some is given to the pan, especially after the roast-
ing has penetrated below the surface layers, and the major
portion is, of course, carried away in the sensible heat of the
current of gas through the pipes and fan. If the temperature
of the gas as it comes directly from the unit were available it
would be interesting to check the heat actually carried away
with the quantity calculated to be lost this way, in the absence
of this data we shall have to be content with the ensuing
figures:
We may consider a charge of one metric ton, which is ap-
proximately one long ton, an amount such as is actually used in
the pans yet tried.
Constituents.
Formula.
% Wet.
% Dry.
Kilos.
SiHcious Ore
SiO.
24
25.5
240
Iron Ore
Fe.O,
22
23.4
220
Pyrite
FeS.
22
23.4
220
Galena
PbS
18
19.2
180
Blende
ZnS
5
h.Z
58
Limestone
CaCO^
3
3.2
30
Moisture
H,0
6
100 100.0 940
The chemical analysis of the charge will show
Insoluble
25.5%
Iron
27.3
Sulphur
16.8
Lead
16.6
Zinc
3.6
Lime
1.8
determined,
Oxygen
7.0
Carbon Dioxide
1.4
100.0
The result of the roasting and sintering cannot be definite-
ly stated as to the exact compounds and the exact amount of
each produced. It is probable that most of the silica will com-
bine with the oxides of lead and iron resulting from the oxida-
tion of their sulphides. Fragments of unchanged silicious ore
are noticeable in the product under certain conditions, it is
equally apparent that in good sinter the silica will be mostly
fused and changed to silicate.
The pyrite we shall assume to be fully decomposed, the
iron oxidized to the ferrous condition and satisfied with silica to
190 THE ARMOUR ENGINEER [Vol. 4, No. 2
I'orm FeSiO^; mis is noi absolutely correct for some pyrite may
escape decomposition, some higher oxide may be formed and
some oxide may not react with silica, however, the assump-
tion is accurate enough for the calculation in hand.
The oxide of iron present in the charge as the iron ore,
Fe^O^, will be more or less fused with the other constituents
it will not take part in any reaction, as here considered, but
being heated and cooling with the remainder of the charge
neither adds nor subtracts in the total heat.
Some small amount of the galena may remain intact dur-
ing the sintering and some of the lead oxide formed may not
react with silica but for the calculation it is assumed that the
reactions are completed. As the product is assumed to contain
only two and nine-tenths per cent sulphur, a considerable por-
tion is present as calcium sulphate, and as zinc sulphide is
oxidized with more difficulty than any of the other sulphides
it seems best to assume that the galena is fully used and that
the lead oxide gets its required amount of silica to form
PbSiO^. We know that the lead silicate forms quickly and
at a lower temperature than any of the other silicates and that
the product contains much lead silicate, the full amount of
lead not appearing in the analysis unless hydrofluoric acid is
used with the other acids in getting the sample into solution.
Although some of the blende will doubtless be acted upon
we deem it best to consider it unchanged; the rapid sintering
and low temperature can hardly favor adequate treatment for
this constituent. It is a fact that if the zinc in the charge runs
as high as five, six or seven per cent the sulphur elimination
is decidedly less than with lower zinc, it has also been noticed
that charges which appear uniformly mixed and of even mois-
ture may not sinter in spots, these patches will be found to
analyze high in zinc, the unchanged crystals of sphalerite ap-
pearing prominently.
As considerable calcium sulphate will certainly be formed
we assume the full amount of carbonate changed to that com-
pound ; the product will, of course, be anhydrous.
We may quite accurately assume that there will be no loss
of metal during the operation ; it is not known how much of
the sulphur is sublimed without oxidation, let us say one-third
of the sulphur of the pyrite acts thus. The amount of sul-
phur trioxide formed is here taken as only enough to form
sulphate with the lime; experiments show that the amount in
the gas must be very small, much less than in the ordinary
up-draft blast-roasting.
A bed of inert material such as limestone, previously
May, 1912] PULSIFER: LEAD ORES 191
made sinter or iron ore will necessarily be added to protect
the grates.
The residue will thus have the following composition :
Insoluble silica 240 Kilos
Ferric oxide 220 "
Ferrous oxide 179.8 "
Lead oxide 167.9 "
Zinc sulphide 50 "
Calcium sulphate 40.8 "
898.5 Kilos
Generation of Heat.
Oxidation of Sulphur to Dioxide
Oxidation of Sulphur to Trioxide
Oxidation of Iron to Ferrous Oxide
Oxidation of Lead to Oxide
Combination of Silica with Lead Oxide
Combination of Silica with Iron Oxide
Union of Calcium Oxide with Sulphur Trioxide
(A) 92.8 (Kgs. S) X 2,164 (Cal. per Kg.)
(B) 9.6 (Kgs. S) X 2,870 (Cal. per Kg.) = 27,571 "
(C) 102.5 (Kgs. Fe) x 1,175 (Cal. per Kg.) = 120,438 "
(D) 155.8 (Kgs. Pb) X 245 (Cal. per Kg.) = 38,171 "
(E) 167.9 (Kgs. PbO) No Data.
(F) 179.8 (Kgs. FeO) x 124 (Cal. per Kg.) = 22,295 "
(G) 18.0 (Kgs. CaO) x 1,676 (Cal. per Kg.) = 30,168 "
439,462 Cal.
Absorption of Heat.
Heat required to decompose Pyrite, (A)
Heat required to decompose Galena, (B)
Heat required to decompose Carbonate (C)
Heat required to vaporize Water (D)
Heat required to vaporize Sulphur, (E)
In the absence of the quantity of heat necessary to decom-
pose Pyrite we will use that required for Ferrous Sulphide, a
quantity probably slightly less.
(A) 220 (Kgs. FeS) x 273 (Cal. per Kg.) = 60,060 Cal.
(B) 18 (Kgs. PbS) X 84.5 (Cal. per Kg.) = 15,210 "
(C) 30 (Kgs. CaCOg) x452 (Cal. perKg.) == 13,560 "
(D) 60 (Kgs. Water) x 606.5 (Cal. per Kg.) = 36,390 "
(E) 39.2 (Kgs. S) X 72 (Cal. per Kg.) = 2,822 "
128,042 Cal.
SiO.,
26.7%
Fe
28.6
Pb
17.3
Zn
3.8
S
2.9
CaO
1.9
(A)
(B)
(C)
(D)
(E)
(F)
xide
(G)
.) =
200,819 Cal.
192 THE ARMOUR ENGINEER [Vol. 4, No. 2
Subtracting 128,042 calories from the total amount of heat
liberated, 439,462 calories, leaves 311,420 calories as the quan-
tity of heat to be radiated away, conducted away, and carried
away in the gases as sensible heat.
The sulphur which is eliminated as dioxide will fill a vol-
ume of some sixty- four cubic meters under standard conditions;
if we suppose four thousand, one hundred cubic meters of gas,
reduced to standard volume, to have been used during a run of
some two hours the average content of the gas in sulphur dioxide
is evidently about one and one-half by volume. The temperature
of the gases will be well below one hundred degrees centigrade
during the first half of the run, toward the last they may get
as hot as four to five hundred degrees centigrade, when the cake
is pretty well finished and the line of fire has entered the pipes.
It would seem that the common method of placing the ex-
haust fan near the unit is ill-advised for both continuous or in-
termittent units. With a continuous machine or several pans
at various stages the temperature of the gases cannot be high
enough to prevent the deposition of the more or less sticky fume,
with a fan for each pan the deposition is excessive during the
first of the run and at the end the fan gets entirely too hot;
the required suction is not more than twelve or fifteen inches
of water and flues can well be constructed so that a large fan
might be placed at some distance thus handling cooler gases and
avoiding collecting the fume which will be settled out where
the individual pipes enter the flue.
The figures indicate plainly the predominating influence of
the sulphur as the source of the heat generation. Practical ex-
periments have determined that with a charge of this general
character twenty-one per cent sulphur is on the upper limit for
the whole mass fuses and stops the reaction, likewise ten per
cent sulphur is too low for the heat liberated is not sufficient
to propagate the combustion zone uniformly to the end.
The roasting and sintering of lead ores is an operation as
necessary in the metallurgy of lead now as it was fifty years
ago; we have tried to record the developments and give some
account of each step in the progress. From every point of view,
efficiency, cost, healthfulness, the conserving of human labor,
time, fuel and metal values and the practical elimination of the
injurious product, sulphur trioxide, developments have been re-
markable and at the present time constitute one of the most ab-
sorbing phases in the metallurgy of lead.
(This paper, as well as a discussion of the other phases of
the Aletallurgy of Lead, by the author, will be found in current
issues of the Salt Lake Mining Review.)
MODERN FARM PROBLEMS AND THE ENGINEER.
BY I. N. BAUGHMAN.*
The greatest source of the food supply of the world is the
soil. With increased population man must find means for both
increasing the productiveness of the soil and for increasing the
producing area. In many parts of the globe only the most in-
tensive methods of agriculture return enough food for the pop-
ulation and often these methods do not suffice to keep hundreds
and thousands of people from starving to death in a single year.
In India and China this condition very often exists and, while
in China and perhaps some parts of India a part of this death
toll is due to insufficient means of transportation, a major part
is due to the impossibility of raising sufficient crops. The
Chinese are noted for their intensive methods and the care with
which they save everything available for fertiHzer. In both
countries grains form the major part of the food, but even with
this condition existing sufficient food is often not available.
In our own land the prices of the principal cereals and of
the live stock used for food has risen year by year, due princi-
pally to the growing demand caused by increased population.
An eminent authority makes the following statement, "Ameri-
can agriculture must develop enormously along new lines to
save the nation from hunger." Our farming area has increased
but demand is overtaking supply so rapidly that the exports of
our agricultural products fell ofif fifty-five million dollars in
twenty years (1899-1909) even with increased prices. Notwith-
standing the great areas of rich virgin soil brought under culti-
vation in the West and Northwest in the last forty years, not-
withstanding the abandonment of worn out lands in the East
and Southeast, and notwithstanding the improvements made in
newer parts of the country in seed, drainage, implements, the
average yield per acre of the principal grain crops of the United
States has not been maintained. The population has during that
time increased more than one hundred per cent. We are one
of the four great agricultural countries (the others being China,
India and Russia) and we must maintain our own population
with our own agriculture as these countries do.
With these facts plainly before us, what means are we using
to help matters and what means will we use in the future? We
will reach the limit of our relief from the sources we are now
*Class of 1910. MarseUles. IHinois.
194 THE ARMOUR ENGINEER [Vol. 4, No. 2
using to help us (viz., extension of farming areas and curtail-
ment of the exportation of foodstuffs) unless other means are
employed. It is conservatively estimated that in thirty years
the population in the United States will reach one hundred and
eighty millions. What then can we do to solve the problem?
Improved methods will help us save ourselves. Scientific
research in our state and government experiment stations are
already bringing forth means to serve better and more profitably
our daily needs ; it is increasing the efficiency of both man and
plant and it is ascertaining processes of farming and fertiliza-
tion which will maintain constant the amounts of the elements
in the soil necessary for plant life. It is even demonstrating
the possibility of increasing the fertility of virgin soils. Through
this research work we are learning the values of irrigation, drain-
age and moisture husbanding, the last being the fundamental
principle of dry farming.
Machinery and the more efficient use of power will probably
be as valuable in helping us as the methods demonstrated by
science pure and simple. This use of power lies more in the
sphere of the engineer than the first methods mentioned, al-
though in these he is involved in many ways.
The power used in agriculture is supplied for the most part
by animals, although other sources are coming into use more and
more. Wherever horses and cattle are kept on a farm the grain
producing area and the available supply of cereals is curtailed.
The increase in the use of mechanical power must necessarily
follow if we are to "keep the wolf from the door." The rea-
sons for this, briefly stated, are: First, production can be in-
creased in amount and quality by confining crop operations to
the time when most favorable conditions prevail ; second, pro-
duction costs can be decreased by the increased "commercial"
efficiency of mechanical over animal power; and third, other
sources than the products of the soil can be used to maintain
mechanical equipment.
Considering the working equipment of the farm as a power
plant we have a very low load factor, the heaviest or peak loads
being high and extending over less than one-eighth of the year.
For the remainder of the time light work or no work at all is
required of the power equipment. Owing to this fact, the great-
est "commercial" or "monetary" efficiency depends on having the
fixed charges on equipment as low as possible. Animal power
is more expensive in first cost, and also in maintenance, than
mechanical power. Engines, tractors and other sources of
mechanical power cost from forty to one hundred and fifty dol-
lars per horsepower, while an equal amount of animal power
May, 1912] BAUGHMAN: FARM PROBLEMS 195-
costs from one hundred and seventy-five dollars to three hundred
dollars. Fuel costs for the latest engine amount to a half cent
per horsepower hour, while that of the animal power is from
four to ten times this amount. Depreciation on mechanical equip-
ment may be somewhat heavier but the risk of loss is greater in
animal equipment. For a long time it will be impossible to ab-
solutely dispense with animal power, if it ever becomes so, but
with improvements in the means of application of mechanical
powers, our dependence on animal pqwer will become less and
less. The direct saving in grain and forage by the displacement
of a single horse amounts to fifty dollars a year and if half of
our work animals could be thus dispensed with, we could save
annually six hundred million dollars, or an amount approximately
equal to our entire wheat crop.
Another very valuable economic factor in the use of power
on the farm is that the cost of manual labor is cut down a great
deal. Two men can operate an engine and eighteen or twenty
plows and control the power of eighty horses. This curtailment
of the use of so much manual labor is important for such labor
is becoming scarcer and scarcer, a fact evidenced by the high
wages paid in the great wheat fields of the Northwest.
Still another very valuable advantage of mechanical power
is that it can be used night and day if necessary, while an animal
equipment cannot do this. The horse tires but the engine can
keep on and on practically without cessation.
In this country, the farms generally consist of one hundred
and sixty acres or more. Such being the case it is possible to
very profitably apply "power" to farm work. Where a smaller
area is handled the fixed charges on mechanical equipment are
too much to warrant their use in the "major" operations unless
companies are formed by a number of farmers to buy and main-
tain the necessary equipment.
There are several ways in which power can be used in crop-
growing: First, it can be used in plowing, tilling and planting;
second, in harvesting and "threshing" (hulling, shelling and
threshing), ^nd third, in transportation to and from the mar-
kets. In irrigation and drainage it also proves to be efficient.
In plowing and tilling there are three methods of using
power: First, by a tractor pulling its plows, harrows, planters,
etc., behind itself securing traction on the ground; second, by
using a stationary prime mover which hauls the implements over
the ground by cables ; and third, by a moving power unit securing
traction by means of a cable fastened at the two ends of the
field. The first method is the mo,st used in this country owing
to the fact that it is the cheapest in first cost. The second method
196 THE ARMOUR ENGINEER [Vol. 4, No. 2
is used abroad and may come into more favor here if electricity
can profitably be distributed over the country. The third method
is not in as great use as either of the others. The disadvantage
of the first method lies in the fact that to secure traction, the
engine or tractor must weigh so much that it will cause injuri-
ous packing of the soil under some conditions. In one type of
tractor this disadvantage is obviated by the use of "caterpillar"
driving wheels which, it is claimed, allows handling of the ma-
chine on wet ground which is impossible for the other type. The
cost of this "caterpillar" tractor is a little too high to be satis-
factory, however. The second method has the disadvantage of
being very heavy and cumbersome and also, as stated above, of
being very much more expensive. The disadvantage of the third
type is that it requires numerous strong cable posts or "dead
men" and involves a great deal of changing of the cable.
As to the operating costs of the tractor, one firm states that
its sixty horsepower "Modern Farm Horse" costs $10.20 per
day to operate in the field as against $19.65 for an equivalent
power furnished by horses. These figures include all items of
expense, are for a ten-hour day and if possible favor the horse-
powered equipment. In harvesting, they figure the cost of horse
as $19.65 and the tractor as $18.43, some handicap being given
the horse in this case also. Another firm figures the cost of
plowing at sixty-six cents per acre, a result not duplicated with
horses or mules. This same firm cites a case where the use of
a tractor cut the cost of production ten cents per bushel on a
wheat crop of twenty bushels per acre in North Dakota, where
it replaced horses.
Besides being more efficient, the tractor allows of deeper
plowing. Dr. S. A. Knapp, of the Department of Agriculture,
who is revolutionizing farm methods in the South, says that out
of a gain of two hundred per cent over an average crop he found
better plowing and pulverizing of the seed bed added one hun-
dred per cent, better cultivation fifty per cent, and better seed
fifty per cent. Under the most widely practiced methods used in
this country the soil is only "scratched" and to achieve and main-
tain the yields which will be required at no very distant time,
of every available bit of tillable land, deep plowing must be prac-
ticed.
After this glimpse at power plowing, tilling and planting, let
us consider the problems confronting us in this phase of agri-
culture.
One of the greatest needs in the way of a power imit, and
one which must be met, is that, for a small tractor, — ^both for
universal use on the small farm and for light work on the large
May, 1912] BAUGHMAN: FARM PROBLEMS 197
one where a greater power unit is also employed. The ultimate
need is for a durable, self-contained unit so light as to be eco-
nomical in the use of power and not to compress the ground,
and capable of filling every use. Owing to the fact that fixed
charges are a large item of expense, it must be low in first cost,
it must be capable of being cheaply handled as far as labor is
concerned, and it must be cheap in repair cost and economical
in fuel consumption. To be cheaply operated from the fuel
standpoint, it must be capable of handling fuels such as kerosene
and distillates. Such engines are even now on the market, but
not in small enough sizes to entirely fill this need. In some
cases steam-driven machinery may be more advantageous than
that with power supplied by an internal combustion engine, but
generally this is not so. The steam-driven outfits require more
manual labor and attendance by men and horses to haul water
and fuel than the internal combustion engines.
In the development of suitable machinery for the use of
power in plowing, tilling, etc., much must be accomplished. This
is also true for the machinery used in the "finishing" processes,
such as harvesting, threshing, etc. The implements and machin-
ery now used are for the most part simply those used with
horses and now changed enough to answer for use with tractors
and other mechanical sources of power. In plowing, new de-
velopments have been made in strictly tractor-operated machin-
ery, and perhaps in pulverizing tools, but it still remains to de-
sign new power seeders, drills and planters, and suitable "hitch-
ing" apparatus.
With regard to harvesting machinery, a great deal has been
accomplished, but the resulting units are only suitable for large
tracts of land. There are numerous outfits in use in the great
wheat fields which "head" the wheat (i. e., cut ofif the heads,
leaving the straw standing), thresh it and deliver the grain to
the wagons as the machine goes across the field. This method
requires a large first cost and although it may be efficient in the
cases in which it is used, it is not applicable to smaller tracts.
Suitable machines may be developed for smaller amounts of
land, which will do this work. In many instances a number of
harvesters of the old horse-drawn type are "hitched" to a tractor,
but this method is not satisfactory, as it involves the use of too
much labor in handling the harvesters and shocking the grain.
A self-contained outfit which will be cheap and small enough for
the smaller tracts of land will obviate the use of a large amount
of this labor and put the process on a more efficient basis.
In harvesting corn, the use of buskers has taken the place of
the farm hand to onlv a small extent. The machines that have
198 THE ARMOUR ENGINEER [Vol. 4, No. 2
been developed, while they work faster than men and require
less labor to operate them, are very heavy and cumbersome.
Owing to the fact that they are driven by "bull wheels," as most
horse-drawn apparatus of a Hke nature is, they must be heavy to
secure sufficient traction to drive the machinery. Being heavy
and somewhat inefficient in the application of the requisite power,
they necessitate the use of too many horses and cannot be op-
erated on soft ground. Lighter, cheaper and more efficient ma-
chines capable of use with power must be developed. In this
connection it may be said that a very hard problem to be solved
will be the development of cultivators and suitable "pulling" ap-
paratus for use in "tending" the corn crop. This is one of the
points where it will be hard to displace the horse.
Until the advent of combined harvesting and threshing units
in smaller sizes, the use of smaller threshers would be advisable
and it is possible that even smaller sizes than those on the mar-
ket at the present time will be developed. In corn shelling, where
it is possible to extend work over a longer time than with thresh-
ing, small sizes are made and they can be profitably used. Power
may be used to a greater extent in graders and grain cleaners
with profit. These machines have not been in general use owing
to the fact that they required power to operate them, but with
the advent of small sizes of both portable and stationary en-
gines, they are a valuable asset in assisting the farmer to get
better quality and better prices for his grain.
The means most generally employed in transporting grain
from the farm to elevators is by horse-drawn vehicles. This is
a very inefficient method, inefficient both in time and labor, as
well as in power. By the use of trucks or wagon trains hauled
by a tractor, less labor is necessary and less time will be wasted.
A number of firms supply suitable trucks and also "tractor" wag-
ons, and their use in various localities has demonstrated their
efficiency. For the most part, the trucks are too high in first
cost and furthermore they have not been worked out with a
view to use in agriculture. For this reason they are not entirely
suitable for an ordinary farmer to operate and handle satisfac-
torily. There are many points about them which must be adapted
to farm use and they must be lighter and cheaper.
In the West, tractors are used a great deal to haul wagon
trains for the transportation of grain. This is very good where
large tracts are handled by one management, but it has not been
used where land is higher in price and smaller tracts are worked
by one individual. There is a possibility of great promise in the
use of such means of transportation in sections where there are
small farms.
May, 1912] BAUGHMAN: FARM PROBLEMS 199
The transportation of grain between the farm and the rail-
roads, as is stated above, is generally very inefficient and better
methods must be used and better roads must be built and main-
tained in many sections. Too much cannot be said on the sub-
ject of good roads and it is hoped that campaigns such as are
being instituted by the National Good Roads Association will
be productive of very far-reaching results. The American
farmer should be alive to the benefits of good roads to himself
and the general public, he should see the waste caused by poor
roads, and be willing and enthusiastic in helping the cause.
After this survey of the conditions as they exist and a few
of the existing problems one can see that the engineer can and
will be a factor. Mechanical power has come to the farm and
displaced animals in countless instances. The large farms are
being broken up into smaller holdings on which the personal in-
terest of the home-loving owner is proving more effective than
the long range management of the wheat baron. A new type of
large farm is made up of small holdings or held by a stock com-
pany, operated by a scientifically-trained manager, and by the
use of mechanical power is multiplying in the West and will
spread over a large part of the country.
The farm has been touched by the soil expert, the chemist,
the botanist and pathologist, the plant and animal breeder, the
economist, and last by the "busmess doctor." Now it is to be
analyzed as an engineering proposition and after all these au-
thorities have laid down the plan, its execution is largely an
engineering problem. The farm of the future will have its suc-
cess measured by the efficiency of the equipment.
Even now courses in agricultural engineering are graduat-
ing men with a knowledge of civil, electrical, chemical, hydraulic,
and most of all, mechanical engineering. The agricultural engi-
neer must have all this and more, he must be an alert, well-
equipped, all-around man. He will install machinery, erect
buildings, irrigate or drain as conditions require and oversee the
maintenance and perhaps the operation of the equipment. He
will be surrounded with all the conveniences and opportunities
for mental development and his life will be in the open. The
man who thus directs the work on the farm and the man who
makes it possible by the designing and building of suitable im-
plements and machinery will be members of an honored profes-
sion, whose careers will be a constant stimulus to breadth of
vision and intellect.
HIGHER HARMONICS IN THREE-PHASE SYSTEMS.
BY E. H. FREEMAN, E. E.*
The presence of higher harmonics in current and voltage
wave-forms is almost universal in alternating current systems.
The order of the harmonics and their relative magnitudes depend
not only upon the physical characteristics of the circuit but also
upon the number of phases. In the single-phase system any odd-
numbered harmonic with any amplitude may exist. In the
three-phase system the third harmonic and its multiples may
be nearly or wholly suppressed or they may be relatively greatly
increased. It is the purpose of the following discussion to point
out some of the things affecting the relative amplitudes of the
harmonics in the three-phase system and to show, by means of
oscillograms, results that have been obtained under certain con-
ditions.
If the winding of an alternating current generator is uni-
formly distributed in a very large number of slots per pole, then
the resultant electromotive-force of fundamental frequency gen-
erated in a group of conductors lying adjacent to each other is
proportional to the number of conductors and to the ratio of
the chord to the arc over which they are distributed. For the
third harmonic electromotive-force of the same group, the same
proposition holds, practically, if the arc is taken three times as
great as it is for the first harmonic, and so on for other har-
monics. Thus, if a group of conductors covers 120 electrical
degrees, as usually measured this becomes 360 electrical degrees
for the third harmonic. Now the resultant electromotive-force
in uniformly distributed conductors covering 360 degrees is zero.
This is also true for any m.ultiple of 360 degrees. It, there-
fore, follows that an electromotive-force which contains a large
number of harmonics may be generated in a single conductor,
but the resultant electromotive-force of a group of these con-
ductors covering 120 electrical degrees will not contain the third
harmonic or any of its multiples.
The type of winding just discussed is practically that found
in synchronous converters and double current generators. When
these machines are connected three-phase, the third harmonic of
electromotive-force and its multiples are almost wholly sup-
*Class of 1902. Professor of Electrical Engineering, Armour Institute of
Tectinology.
May, 1912] FREEMAN: HIGHER HARMONICS
201
pressed, regardless of the flux distribution in the air gap. It is,
of course, impossible with a slotted armature, for the groups
of conductors to fully cover 120 electrical degrees, so that the
third harmonic is not quite reduced to zero, but it becomes so
small that it is practically negligible.
For three-phase alternating-current generators, the usual
type of winding for one phase consists of a group of conductors
covering an arc of about 60 degrees connected in series with
another group covering a like angle, the centers of the groups
being 180 degrees apart. Now a distribution of conductors over
an arc of 60 degrees does not suppress any of the higher har-
Fig. 1.
A^, Bj, Cj — Fundamental Eleotroniotive-force.s of Different Phases.
A3, B3! C3— Third Harmouie Electromotive-forces of Different Phases.
R_jResultant Fundamental Electromotive-force Acting Around Delta.
R^ Resultant Third Harmonic Electromotive-force Acting Around Delta.
monies to the same extent that a distribution through 120 de-
gress does. An arc of 60 degrees for the first harmonic is one
of 180 degrees for the third harmonic, 300 degrees for the fifth
harmonic and so on. The resultant third harmonic electromotive-
force is, therefore, not zero, and the same condition holds for
all other odd harmonics, when any of these harmonics are gen-
erated in the individual conductors.
It should also be noted that the connecting of a second
group of conductors in series with the first does not change the
relative magnitudes of the various harmonic components of the
resultant electromotive-force when centers of the groups are 180
202
THE ARMOUR ENGINEER
[Vol. 4, No. 2
degrees apart. The resultant first harmonic electromotive-forces
of the groups will be in phase with each other and add directly.
The higher harmonics in each group will have the same phase
relations to the fundamental electromotive-forces, hence they
will add directly. A series grouping of such a winding, therefore,
does not affect relative values, and the elimination of higher
harmonics in any one phase winding must be brought about by
a proper distribution of the flux in the machine.
However, an entirely different condition exists when three
such windings as discussed above are connected to form a delta
or a star. With the delta connection the first harmonic elec-
Fig. 2.
Aj, Bj — Fundameutal Klectroiuotive-forces of Two Phase-AVindings.
R^ — Resultant Fundameutal Electroiuotive-foree When Windings Are
Connected in Star.
Rg — Resultant Third Harmonic Electromotive-force AVhen AVindings Are
Connected in Star.
tromotive-forces differ in phase by 120 degrees and the result-
ant, acting around the delta, is zero. But the third harmonic
electromotive-forces of the separate phase-windings are in phase
with each other (see Fig. 1) and therefore give a resultant equal
to three times one component. The same condition exists for all
multiples of the third. All other higher harmonics have, like
the first, a resultant of zero with reference to a circuit around
the delta.
The effect of the third harmonic electromotive-force and
its multiples is to produce a current around the delta, and the
impedance drop of this current will equal the sum of the gener-
May, 1912] FREEMAN: HIGHER HARMONICS
203
ated electromotive-forces producing it. The result is that the
line voltage is almost wholly free from these harmonics, but it
may contain others. This current produces an extra copper loss
and is objectionable for this reason.
When the windings are star-connected, the first harmonic
electromotive-forces of two phase-windings add at 60 degrees
phase difference to give the line voltage. The third harmonic
electromotive-forces, however, if present, add at 180 degrees
phase difference, giving a resultant of zero (see Fig. 2). The
same condition exists for all multiples of the third harmonic.
The line voltage, therefore, cannot contain any of these.
It is important to observe that the third harmonics in the
three windings of the star connection are in phase with each
other when considered with reference to the neutral, hence
these electromotive-forces tend to send simultaneously currents
out on the lines, and back on the neutral and vice-versa, alter-
nately. If the neutral of a generator is connected to the neutral
of a balanced star-connected load, then a current of triple fre-
quency and multiples thereof may exist in the neutral connection
and the main lines. Each line will carry one-third of the current
that exists in the neutral connection. If the neutral connection
is opened this breaks the circuit for this current and a voltage
204
THE ARMOUR ENGINEER
[Vol. 4, No. 2
will appear across the break which contains the same harmonics
as did the neutral current.
A condition differing in some respects from the above exists
when two star-connected alternators are operated in parallel. If
the electromotive forces of the two machines are equal in
magnitude and identical in wave-form and directly opposed
to each other, no cross current will exist between them.
If the electromotive-forces are equal and identical in wave-
form but are not directly opposed, then a current will
exist. If there is no connection between the neutrals, this
current cannot contain a third harmonic or any multiple thereof,
but it may contain any other higher harmonics. A voltage will
exist between the neutrals of the generators which may contain
the third harmonic and any of its multiples, but no others so
long as the impedances of the various phases and the connec-
tions thereto are balanced.
It is, of course, the usual condition in parallel operation
for slight phase differences to exist. Exact equality of electro-
motive-forces is seldom obtained, and identical wave-forms rnay
not be found even in machines constructed as near as practical
on the same dimensions. All of these causes contribute to the
production of a current between the machines. If the neutrals
May, 1912] FREEMAN: HIGHER HARMONICS
205
are not connected, this current will not contain a third har-
monic or any of its multiples, but a voltage will exist between
the neutrals which contains these harmonics only.
The oscillogram of Fig. 3 illustrates the condition just
stated. Two small star-connected alternators of the same size
and make were excited to give practically the same voltage and
then connected in parallel without a load. The larger wave in
the figure is that of the bus-bar voltage ; the smaller, the voltage
between the neutrals. In this case the third harmonic in the
voltage wave is relatively small and the ninth is more prominent.
When the excitations of the two machines differ consider-
To Load
Fig. 5
ably, currents exist in the machines- which raise the voltage of
the under-excited and lower the voltage of the over-excited
machine. With no connection between neutrals, a voltage exists
between these points containing the same harmonics as before
but of relatively different values. This is shown in the oscillo-
gram of Fig. 4. In this case the voltage of one machine was
about ten per cent below and the other about ten per cent above
normal before they were connected in parallel. They were
operated without load. The larger wave is the bus-bar voltage
and the smaller one that between neutrals. Here the third
harmonic is more prominent than in Fig. 3 and a small value of
206 THE ARMOUR ENGINEER [Vol. 4, No. 2
fundamental voltage also shows between neutrals. The latter
is probably due to inequality in the connections between the
generators.
It should be stated that the ordinates to the neutral voltage
wave are greatly magnified over those for the bus-bar voltage.
The effective value of the neutral voltage in this case was about
one-half of one per cent of the line voltage. While this appears
to be a negligible value, it was found with these machines that
the current in a neutral connection might amount to five per
cent of the full-load line current. The oscillogram of Fig. 5
shows such a case. A balanced non-inductive load was con-
Fig. 6.
nected to the generators so that each machine was carrying
about one-third of its full-load current.
The neutral current-wave shown had an effective value
of eighteen and one-half per cent of the line current from each
generator. A reason for the current being relatively greater
than the electromotive-force is that, on account of the neutral
current dividing equally among the three windings of the ma-
chine and these three components being in phase, the self-induc-
tion was quite small. The resistance of the circuit and the leak-
age reactance of the windings are the only things opposing
the current.
May, 1912] FREEMAN: HIGHER Hx\RMONICS
207
The small value of the first harmonic appearing in the
neutral current is probably due to unequal impedances in the
lines connecting the generators in parallel.
With machines of different sizes and different makes oper-
ating in parallel, it is very probable that greater effects will
exist than have been cited above. As an illustration, two alter-
nators, one having twice the rating of the other and being
made by different manufacturers were experimented upon. The
larger machine generated a higher voltage and this was stepped
down through three transformers star-connected both on the
primary and on the secondary sides, before the machines were
connected in parallel. The neutral of the larger generator was
connected to the neutral of the transformers on the primary
side so that the wave forms of voltage in the secondaries were
practically identical with those of the generator. Fig. 6a shows
the line voltage and line-to-neutral voltage waves of the small-
er machine and Fig. 6b, the same waves for the larger machine,
when they were running separately and without load.
When these machines were connected in parallel and sup-
plied power to a small lamp-load equal to about ten per cent
of their combined rating the wave forms shown in Fig. 7 were
obtained. In this case there were no neutral connections to the
208
THE ARMOUR ENGINEER
[Vol. 4, No. 2
machine or to the load. The voltage between neutrals had an
effective value of nine per cent of the line voltage. As shown,
it contains prominent third and ninth harmonics.
With the neutral of the smaller generator connected to the
secondary neutral of the transformer, there was a current in this
connection having an effective value of over ten per cent of the
full-load current of the smaller machine. This is a much great-
er current than was found when the two machines of the same
size were operating together. The wave shape of this neutral
current is shown in Fig. 8. It contains, as did the voltage wave,
a third and a ninth harmonic. These are of relatively different
Fig 8.
amplitudes and different phases from the same components in
the voltage wave.
The load current has remained practically unchanged,
showing that the neutral current exists only in the generator
and transformer windings and in the connections between them.
In this case, it not only serves no useful purpose, but increases
the copper loss.
When a single-phase generator giving approximately a sine
wave of voltage is connected to a transformer without load, the
exciting current, due to hysteresis and varying permeability, is
non-harmonic, and contains, among others, a component of
May, 1912] FREEMAN: HIGHER HARMONICS
209
triple frequency. When a three-phase generator is connected
to three identical single-phase transformers, no triple-frequency
current can exist in the lines connecting the generator with the
transformers unless there is a neutral connection. Of course, no
such connection can be made with delta grouping of the wind-
ings, though triple-frequency currents may exist in the delta
apart from other connections.
To show the effects with transformers, one of the ma-
chines mentioned in the previous experiments was connected
to three single-phase shell-type transformers. The generator
and primaries of the transformers were star-connected and the
Fig.
secondaries were delta-connected. Fig. 9 shows the generator
voltage wave from line to neutral, and the line current and
voltage between the neutral of the generator and the neutral
of the transformer connections. It is apparent that the line
current departs but slightly from sine form. Its value was
about three per cent of the full-load current. The voltage be-
tween neutrals contains a small third harmonic with a rather
prorninent ninth. It should be added that the secondary delta
carried an exciting current composed of a third harmonic and
its multiples, but unfortunately no oscillogram is available to
show it.
210
THE ARMOUR ENGINEER
[Vol. 4, No. 2
When the neutral of the generator was connected to the
neutral of the transformer group the circuit was closed for any
triple-frequency currents or multiples thereof. The result is
shown in Fig. 10. The neutral current wave here shows a third
and a ninth harmonic which combined with the line current
previously shown and gave a greatly distorted wave. The ex-
citing current in the main lines was found to have increased in
value from about three per cent to five and three-tenths per
cent of full-load current. The generator now supplied all of
the exciting current directly, while in the previous case a part
composed of the third harmonics and some of its multiples
existed in the secondary delta. The neutral current was about
fourteen per cent of full-load line current.
If the primaries and the secondaries of the transformers
are star-connected and no neutral connection is made to either,
triple-frequency currents cannot exist in any part of the cir-
cuit, and in their place triple-frequency electromotive-forces
\w\\\ be found. To demonstrate this, the generator whose volt-
age wave-forms are shown in Fig. 6b was connected to three
single-phase transformers with primaries and secondaries in
star. Fig. 11 shows the line-to-neutral voltage at the trans-
formers, the line current and the voltage between neutrals.
May, 1912] FREEMAN: HIGHER HARMONICS
The line-to-neutral voltage, which in this case is the voltage
across one transformer, shows the presence of a prominent
third harmonic component. This is due to the fact that the
exciting current, in which a third harmonic cannot exist, ap-
proximates a sine wave and therefore produces, on account of
varying permeability, a flat-topped flux wave in the transformer.
Such a flux wave contains, among others, a third harmonic and
produces a peaked electromotive-force wave containing a third
harmonic electromotive-force relatively three times greater than
the third in the flux wave.
The voltage between neutrals is largely due to the third
Fig-. 11.
harmonic generated in the windings of the transformers. In
this case, the effective value of the voltage between neutrals
was thirty-eight and one-half per cent of the line-to-neutral
voltage. _ This relative value depends upon the degree of sat-
uration in the magnetic circuit and upon secondary connections.
For example, when the line-to-neutral voltage was increased
eight per cent, the voltage between neutrals increased twenty
per cent, showing clearly the effect of higher flux density in
the transformers. With the secondaries in delta, the neutral
voltage was but five and one-half per cent of that between line
and neutral.
212
THE ARMOUR ENGINEER
[Vol. 4, No. 2
The high value of neutral voltage (secondaries in star) does
not produce as large a neutral current as might be expected,
when the two neutrals are connected through a low resistance.
With such a connection the line-to-neutral voltage of the gen-
erator is impressed across each transformer ana the form of
the current wave will depend upon the form of the voltage wave
of the generator. With the machine used in this case, the neu-
tral current amounted to about twenty-five per cent of the ex-
citing current in the main lines and to a very much smaller
percentage of the full-load current of the transformers. The
wave forms found for this connection are shown in Fig. 12,
that of the neutral current being drawn relatively larger than
the line current.
The conditions existing with three-phase core-type trans-
formers is decidedly different from that with single-phase
transformers. The three-phase core-type transformer has mag-
netic circuits corresponding in some respects to the electric cir-
cuits in a three-phase three-wire system. It is impossible for
a triple-frequency current to exist in the main lines of such a
system, and it is also impossible, barring leakage, for a triple-
frequency component of the flux to exist in the cores of such a
transformer. When an alternator is connected to three single-
May, 1912] FREEMAN: HIGHER HARMONICS
213
phase transformers, the primaries being in star and the sec-
ondaries in star, or separate from each other, it has been shown
that the voltage wave from hne to transformer neutral con-
tains a third harmonic. These third harmonic electromotive-
forces in the transformers are in phase with each other, and,
therefore, the third harmonic fluxes which produce them are
in phase. In the single-phase transformers these fluxes exist
independently in separate circuits. In the three-phase core-type
transformer, however, the flux of one phase must pass through
the cores of the other two phases, and this works nicely when
the fluxes are 120 time-degrees apart. But any third harmonics
Fig. 13.
of flux, being in phase with each other, would tend to exist
along the cores in the same direction simultaneously. The re-
sultant of these would, therefore, be zero. It follows from this
that no third harmonic electromotive-force will be generated in
any of the transformer windings, and the same condition holds
for all multiples of the third.
An illustration of the actions just discussed is given in Fig.
13. The largest wave is the line voltage of a star-connected
alternator ; the medium wave, the voltage from line to neutral ;
and the smallest wave, the voltage between the neutral of the
generator and the neutral of a three-phase core-type transform-
214 THE ARMOUR ENGINEER [Vol. 4, No. 2
er, secondaries not connected. This neutral voltage contains a
small fundamental component probably due to an unbalanced
condition in some part of the circuit. It also contains a dis-
tinct twenty-first harmonic along with a prominent third. The
twenty-first harmonic is easily seen in the line-to-neutral volt-
age of the generator and the straight sides of this wave also
suggest the presence of a small third harmonic. It, therefore,
seems reasonable to suppose that practically all of the voltage
between neutrals is due to the generator and none to the trans-
former. The ordinates to this wave are increased relatively
over those for the other wave. Its effective value was two
and four-tenths of the line-to-neutral voltage, a value very
much less than that obtained with single-phase transformers
under the same conditions. It was also found, on connecting
the neutrals, that the neutral current amounted to about four-
hundredths of an ampere, a striking difference from that in the
case of the single-phase transformers.
The foregoing discussion has pointed out and illustrated
only a part of the phenomena connected with higher harmonics
in three-phase systems, but sufficient has been given to indicate
the peculiar prominence or lack of it, as the case may be, of the
third harmonic and its multiples. When these are present to a
serious extent, a study of the conditions which produce them
will show what must be done to lessen their values or eliminate
them entirely.
SOME CREOSOTED WOOD BLOCK PAVEMENTS IN
CHICAGO.
BY WM. F. HARVEY, C. E.*
The experience the City of Chicago has had with creosoted
wood block pavement extends over a period of thirteen years.
The early pavements were laid on streets in different parts of
the city, thus trying out this type of pavement under various
traffic conditions. The pavements have given such general sat-
isfaction that creosoted wood block has steadily gained in popu-
larity as a paving material for streets and alleys where a noise-
less pavement is required. The object of this paper is to give a
brief history of some of the first creosoted wood block pave-
ments laid in this city, together with extracts of the specifica-
tions governing the kind of preservative used in the manufac-
ture of the blocks and the manner of laying them.
In 1899 the two roadways, each twenty feet in width, of the
Rush street bridge were paved with wood block. It has been
stated that these roadways carried as heavy traffic as that of any
other street in the city at that time. The blocks were of long
leaf yellow pine, rectangular in shape and cut to dimensions
of four inches in depth by four inches in width and about eight
inches in length. Those laid on the east roadway were impreg-
nated with sixteen pounds of the preservative per cubic
foot of timber. The west roadway was paved with blocks of
the same timber, cut to like dimensions, but having received no
preservative treatment. The life of the untreated blocks was
but three years when they were removed. The creosoted blocks
were removed at the end of nine years, because the untreated
planks on which they were laid had to be replaced because of
decay. The blocks, when removed, were in good condition,
showing from one-eighth inch to one-half inch wear, but no
signs of decay, and would no doubt have given good service for
several years longer had the foundation been permanent.
A section of the roadway of Michigan avenue between
Congress street and Van Buren street was paved in 1900 with
creosoted long-leaf yellow pine blocks, five inches in depth,
three and three-quarters inches in width, and nine inches in
length. The blocks were manufactured by the Repubhc Creo-
soting Company and given to the South Park Commissioners
♦Class of 1905. Assistant Engineer, Board of Local Improvements, City
of Chicago.
216 THE ARMOUR ENGINEER [Vol. 4, No. 2
to lay for a sample pavement. The blocks had been impregnated
with sixteen pounds of creosote oil per cubic foot of timber.
They were laid with the direction of the fibre vertical, on top
of a dry cement grout cushion which was evenly distributed
over a concrete foundation. Adjoining it an area of asphalt
blocks was laid at the same time. The traffic on the roadway
was that of light vehicles and motor cars. At the end of five
years the asphalt block wearing surface had worn out and was
replaced with creosoted wood blocks to conform with the wood
block pavement in place. These pavements remained intact
until 1909, when the roadway was widened and the pavement
removed to allow a recrowning of the roadway. The blocks,
after being in the pavement nine years, were found to be in a
sound condition, with no signs of decay and showing an average
of but one-eighth inch wear.
West Taylor street, between South Canal street and Blue
Island avenue, was paved with this kind of material in 1904.
The blocks were of long-leaf yellow pine cut to dimensions of
four inches by four inches by five to ten inches, and were im-
pregnated with twelve pounds of creosote oil per cubic foot of
timber. They were laid on a one-inch sand cushion on a six-inch
Portland cement concrete foundation. The width of the road-
way is thirty-eight feet, with a double-track street railway down
the center. The traffic is heavy and confined to a narrow width
of pavement. This pavement has been in continuous use since
laid, and is in excellent condition at the present time. In the
winter of 1910, after six years of wear, a block was taken from
the roadway half way between the car tracks and the curb, and
a measurement of its depth showed an average of one-eighth
inch wear. The fibres of the block were uninjured, it being
sound and showing no signs of decay.
Astor street, from Burton place to North avenue, was paved
with creosoted wood blocks in 1905. The blocks were of long-
leaf yellow pine treated with sixteen pounds of creosote oil per
cubic feet of timber. They were three and one-half inches in
depth and were laid on a one-inch sand cushion on six inches
of Portland cement concrete. The pavement is still in use and
in good condition.
Examination of the four pavements mentioned above tends
to show that a well-constructed pavement of creosoted wood
blocks, properly manufactured, will give good service under
(1) heavy teaming traffic, (2) boulevard traffic, (3) moderate-
ly heavy business traffic, and (4) light residence traffic. Where
noise is an important factor, as in the "loop" district, creosoted
wood block is rapidly taking the place of granite block as a
May, 1912] HARVEY: WOOD BLOCK PAVING 217
paving material. Following is a list of streets in the "loop" that
have been paved with wood block:
street. From To Paved
Monroe St Clark St Dearborn St 1907
Adams St Clark St Dearborn St 1908
Quincy St Dearborn St State St 1909
Adams St State St Wabash Av 1909
Dearborn St. (w. side) . . Alley south of Monroe. .Adams St 1909
Dearborn St. (e. side).. .Alley south of Monroe. .Jackson Blvd 1909
Clark St. (e. side) Adams St Jackson Blvd 1909
Adams St Market St State St 1910
(Except Dearborn Street to Clark Street)
La Salle St Madison St .Jackson Blvd 1910
Madison St Market St State St 1910
Monroe St Dearborn St Michig-an Av 1910
Randolph Street System:
Randolph St State St Michigan Av 1910
W^abash Av Randolph St Washing-ton St 1910
City Hall System:
Randolph St Clark St La Salle St 1910
Clark St Madison St Lake St 1910
La Salle St Madison St Randolph St 1910
Washing-ton St La Salle St Clark St 1910
Monroe St Clark St La Salle St 1911
Randolph St State St Clark St 1911
Randolph St Fifth Av La Salle St 1911
Dearborn St Monroe St Alley south 1911
Clark St Madison St Van Buren St 1911
Each of the pavements was laid on a one-inch sand cushion
and Portland cement concrete foundation. The blocks were of
long-leaf yellow pine cut to dimensions of four inches by four
inches by six inches to twelve inches or four inches by three and
three-quarters inches by five inches to ten inches, laid with the
grain of the wood vertical, making a wearing surface four inches
thick. The blocks had previously received a preservative treat-
ment of either twenty pounds or sixteen pounds of creosote oil
per cubic foot of timber. Expansion joints placed at proper
intervals and filled with paving pitch ("the direct resuh of the
distillation of straight-run coal tar") were constructed to take
care of the expansion and contraction of the blocks due to mois-
ture absorption and changes in temperature. Coal-tar paving
pitch was used as a filling for the interstices between the blocks
in many of the contracts, while in a few cases dry sand was used
as a filler in place of the pitch. The writer does not favor sand
as a filler for wood block pavements.
Fig. 1 is a photograph of LaSalle street in the block
between Madison street and Monroe street, showing the method
of laying the blocks on the one-inch sand cushion which has
been evenly distributed over the eight-inch concrete founda-
tion. The one-inch wooden strips along the curbs are with-
drawn after the wearing surface has been rolled, and the exist-
ing joints filled with the paving pitch used for expansion joints.
Transverse expansion joints were placed at intervals of twenty-
five feet in the pavement.
218
THE ARMOUR ENGINEER
[Vol. 4, No. 2
Fig. 2 shows the wearing surface in place and being
rolled. The final inspection of the blocks is made at this time
and imperfect ones removed and replaced with perfect blocks.
After the pavement is rolled the longitudinal and transverse
expansion joints are poured and the filler applied to the surface
and brushed into the interstices between the blocks. The sur-
face of the pavement is then covered to a depth of one-fourth
inch with screened hot torpedo sand, and rolled with a light
steam roller before the street is thrown open to traffic.
LJ
f
^
Fig. 1. LaSalle Street, Bet«een Madison and Monroe Streets.
Fig. 3 shows the pavement completed on LaSalle street
between Washington street and Randolph street.
Fig. 4 shows the laying of the blocks on the east side of
Clark street between Washington street and Randolph street.
The following extracts from the current specifications will
govern the cushions and fillers to be used, and the method of
laying the blocks as well as the creosote oil preservative used in
the manufacture of the paving blocks.
Cushion.
Upon the concrete foundation shall be spread a layer of
torpedo sand, free from loam and dirt, in such quantity as to
insure, when compacted, a uniform thickness of ( . . )
May, 1912]
HARVEY: WOOD BLOCK PAVING
219
inches. In surfacing said layers of sand the contractor shall
use such guides and templets as the engineer may direct.
Upon the concrete foundation shall be spread a layer com-
posed of one (1) part of Portland cement to four (4) parts
torpedo sand, thoroughly mixed and dry, and in sufficient
quantity to insure, when compacted, a uniform thickness of
( . . ) inches. In surfacing said layer the contractor
shall use such guides and templets as the engineer may direct.
Immediately before laying the blocks the mixture shall be wetted
Fig. a. LaSalle Street, UetwetMi Madison and Monroe Streets.
by means of a rose-head sprinkler with just sufficient water to
partially cake it.
Laying.
The blocks shall be laid in parallel courses across the road-
way at an angle of approximately ( . . ) degrees
from the center line thereof, except at the intersections of all
alleys, where they shall be laid at right angles with the center
lines thereof. On intersections and junctions of lateral streets,
the blocks shall be laid at an angle of forty-five (45) degrees
with the line of the street, unless otherwise ordered by the engi-
neer. The blocks shall be laid with the fibre of the wood running
in the direction of the depth. Gutters shall be constructed as
220
THE ARMOUR ENGINEER
[Vol. 4, No. 2
directed by the engineer. The courses shall break joints alter-
nately by a lap of not less than two (2) inches and the blocks
shall be driven together except where joints for expansion are
constructed as follows : On each side of the roadway a longi-
tudinal joint shall be formed by placing a one and one-half
{lyi) inch board on edge against the curb. The blocks shall be
firmly laid against said boards. The boards shall remain in
place until the blocks are rolled, and immediately preceding the
application of the filler as hereinafter specified they shall be
carefully removed without disturbing the adjacent blocks.
';fe»
w
I'm
Fig. 3. LaSaUe Street. Between AVashinston aud Raudolph Streets.
The blocks, when set, shall be rolled with a steam roller
weighing not less than five (5) tons, until firmly bedded and
brought to a uniformly even surface. After rolling, all imper-
fect blocks shall be removed and replaced by perfect blocks.
Broken blocks shall not be used except to break joints in start-
ing courses and in making closures. If the blocks that have been
laid should become wet before the filler is applied, they must be
taken up and reset at the contractor's expense, if the engineer
so directs. In no case will teams be allowed on the work before
the wearing surface is completed.
Asphaltic Filler.
After rolling, the surface of the pavement shall be swept
clean and the joints between the blocks and expansion joints
shall be filled with an asphaltic filler which shall be free from
May, 1912] HARVEY: WOOD BLOCK PAVING
221
222 THE ARMOUR ENGINEER [Vol. 4, No. 2
water, coal-tar pitch, or any product of coal or water gas tar.
It shall adhere firmly to the blocks, be pliable at all climatic
conditions to which it will be subjected, and conform to the
following requirements :
It shall have a specific gravity of not less than nine hun--
dred and sixty-five thousandths (0.965) at seventy-seven {77)
degrees Fahrenheit.
It shall have a melting point of not less than one hundred
and ten (110) and not more than one hundred and sixty (160)
degrees Fahrenheit.
It shall have a penetration of not less than twenty (20)
nor more than fifty (50).
The bitumen of the asphaltic filler shall be soluble in carbon
tetra chloride to the extent of at least ninety-eight and one-half
(983^) per cent.
The asphaltic filler shall be heated to a temperature of not
less than two hundred and eighty (280) degrees nor more than
three hundred and fifty (350) degrees Fahrenheit, and shall be
applied in such a manner that all spaces between the blocks will
be completely filled, the temperature of heating to be varied
within these limits according to the nature of the asphaltic filler
used and at the discretion of the Board of Local Improvements.
In applying the asphaltic filler care must be taken to use the
least amount necessary to properly fill the joints and hold the
top dressing. The blocks must be dry at the time of the appli-
cation of the filler.
The contractor shall provide the Board of Local Improve-
ments with a duplicate delivery ticket for each and every con-
signment of asphaltic filler delivered on the work. This ticket
must be signed by the consignor and be of a form approved by
the Board of Local Improvements.
Pitch Filler.
After rolling, the surface of the pavement shall be cleaned
and the joints between the blocks and expansion joints shall be
filled with a "straight-run" paving pitch obtained from gashouse
tar. No pitch from coke oven tar shall be used. It shall be of
such quality and consistency as will be approved by the Board
of Local Improvements. The pitch shall contain not less than
twenty-eight (28) per cent nor more than thirty-five (35) per
cent of free carbon, and shall have a melting point at a tem-
perature of not less than one hundred and forty-five (145) de-
grees and not more than one hundred and fifty-five (155) degrees
Fahrenheit. The pitch must be used at a temperature of not
less than three hundred (300) degrees and not more than three
hundred and fifty (350) degrees Fahrenheit. In applying the
May, 1912] HARVEY: WOOD BLOCK PAVING 223
pitch, care must be taken to use the least amount necessary to
properly fill the joints and hold the top dressing.
The contractor shall provide the Board of Local Improve-
ments with a duplicate delivery ticket for each and every load
or tank of paving pitch delivered on the work. This ticket must
be signed by the consignor and be of a form approved by the
Board of Local Improvements.
Cement Grout Filler.
After rolling, the surface of the pavement shall be swept
clean and the joints, except as hereinafter provided, between
the blocks shall be filled with a cement grout filler composed of
equal parts by volume of clean, sharp, dry sand and Portland
cement, the same to be thoroughly mixed dry, after which water
shall be added, forming a liquid of the consistency of thin cream.
From the time the water is added until the grout is floated into
the joints of the pavement, the mixture must be kept in con-
stant motion, and immediately after applying to the pavement
it shall be thoroughly swept into all the joints. The grout or
filler shall be applied in two or more courses. The first course
shall fill the interstices between the blocks to within two and
one-half (2^) inches of the top, the same to be left undis-
turbed and sufficient time allowed to elapse for the first appli-
cation to stiffen. The following courses or applications shall be
mixed in like manner, except that the mixture shall be slightly
thicker than that of the first course. To avoid a possibility of
the grout thickening at any point, water shall be applied ahead
of the sweeping by spraying.
The expansion joints and joints between the blocks in a
space of two (2) feet in width adjacent to the gutters and
around all covers to sub-surface improvements, shall be filled
with a "straight-run" paving pitch obtained from gas house tar.
No pitch from coke oven tar shall be used. It shall be of such
quality and consistency as will be approved by the Board of
Local Improvements. The pitch must be used at a temperature
of not less than three hundred (300) degrees and not more than
three hundred and fifty (350) degrees Fahrenheit. The pitch
shall contain not less than twenty-eight (28) per cent nor more
than thirty-five (35) per cent of free carbon, and shall have a
melting point at a temperature of not less than one hundred
and forty-five (145) degrees and not more than one hundred
and fifty-five (155) degrees Fahrenheit.
Sand Filler.
After rolling, the surface of the pavement shall be swept
clean and the joints, except as hereinafter provided, between
224 THE ARMOUR ENGINEER [Vol. 4. No. 2
the blocks shall be filled with clean, warm, fine sand, which shall
be swept into the joints until the same are completely filled.
The expansion joints and joints between the blocks in a
space of two (2) feet in width adjacent to the gutters and
around all covers to sub-surface improvements shall be filled
with a "straight-run" paving pitch obtained from gas house tar.
No pitch from coke oven tar shall be used. It shall be of such
quality and consistency as will be approved by the Board of
Local Improvements. The pitch must be used at a tempera-
ture of not less than three hundred (300) degrees and not more
than three hundred and fifty (350) degrees Fahrenheit. The
pitch shall contain not less than twenty-eight (28) per cent
nor more than thirty-five (35) per cent of free carbon and shall
have a melting point at a temperature of not less than one hun-
dred and forty-five (145) degrees and not more than one hun-
dred and fifty-five (155) degrees Fahrenheit.
Top Dressing.
Immediately after the filling of the joints, the surface of
the pavement shall be covered to a depth of one-quarter (^)
inch with screened, hot, torpedo sand.
1. The oil shall be a distillate obtained wholly from coal tar.
2. It is required by this specification that the oil used shall
be wholly a distillate oil obtained only by distillation from coal
tar. No other material of any kind shall be mixed with it.
3. The oil shall contain not more than one (1) per cent
of matter insoluble in hot benzol and chloroform.
4. Its specific gravity at twenty-five (25) degrees Centi-
grade shall not be less than one and eight-hundredths (1.08)
and not more than one and twelve-hundredths (1.12).
5. The oil shall be subject to a distilling test, as follows :
The apparatus for distilling the creosote must consist of
a stoppered glass retort having a capacity, as nearly as can be
obtained, of eight (8) ounces up to the bend of the neck, when
the bottom of the retort and the mouth of the ofif-take are in
the same plane. The bulb of the thermometer shall be placed
one-half (^) inch above the liquid in the retort at the beginning
of the distillation, and this position must be maintained through-
out the operation. The condensing tube shall be attached to the
retort by a tight cork joint. The distance between the ther-
mometer and the end of the condensing tube shall be twenty-
two (22) inches, and during the process of the distillation the
tube may be heated to prevent the congealing of the distillates.
The bulb of the retort and at least two (2) inches of the neck
must be covered with a shield of heavy asbestos paper during
May, 1912] HARVEY: WOOD BLOCK PAVING 225
the entire process of distillation, so as to prevent heat radia-
tion, and between the bottom of the retort and the flame of the
lamp or burner two (2) sheets of wire gauze each twenty (20)
mesh fine and at least six (6) inches square must be placed.
The flame must be protected against air currents.
The distillation shall be continuous and uniform, the heat
being applied gradually. It shall be at a rate approximating
one (1) drop per second, and shall take from thirty (30) to
forty (40) minutes after the first drop of distillate passes into
the receiving vessel. The distillates shall be collected in weighed
bottles and all percentages determined by weight in comparison
with dry oil. When one hundred (100) grams of the oil are
placed in the retort and subjected to the above test, the amount
of distillate shall. not exceed the following:
Up to 150 degrees Centigrade, 2 per cent.
Up to 210 degrees Centigrade, 10 per cent.
Up to 235 degrees Centigrade, 20 per cent.
Up to 315 degrees Centigrade, 40 per cent.
The distillation of the oil shall be carried to three hundred
and fifty-five (355 ) degrees Centigrade. The residue thus ob-
tained, when cooled to fifteen (15) degrees Centigrade, shall not
be brittle, but shall be of a soft waxy-like nature so that it can
be readily indented with the finger. When a small portion of
this residue is placed on white filter paper and warmed, the oil
spot produced, when viewed by transmitted light, shall appear
of an amber color.
The contractor shall deliver to the Board of Local Im-
provements an affidavit from the individual manufacturing the
blocks (if manufactured by an individual), from the managing
officer of the corporation manufacturing the blocks (if manu-
factured by a corporation), and by an active member of the
firm manufacturing the blocks (if manufactured by a firm),
setting forth that all oil used for treating the blocks for this
contract is a distillate oil obtained wholly and entirely by dis-
tillation from coal tar and that it is free from. any adulteration.
There have been laid to date in this city approximately
twenty-four miles of street and two miles of alley pavements
of creosoted wood blocks.
THE AUTOMATIC CONTROL OF FIRE.
BY FITZHUGH TAYLOR.*
The problems which confront the fire protection engineer in
the practice of his profession may be classified into two general
groups, more or less allied in certain individual instances, and yet
as a rule quite distinct one from the other. The first comprises
those relative to prevention of fire, and involves observance
of proper methods of construction of buildings, avoidance or re-
moval of unnecessary fire hazards and intelligent segregation and
safeguarding of hazards which are necessary to the conduct of
certain industries or inseparable from them. The second includes
problems bearing upon control and extinction of fires, and covers
design, installation and maintenance of apparatus and equipment
employed to extinguish fire or oppose its progress.
In dealing with problems of the second group it is in cer-
tain instances possible to provide only manually-operated pro-
tective equipment to be handled during fires by employes of
property owners, municipal firemen or both, but in many cases
it is now practicable to take such measures as will furnish a
reasonable degree of assurance that fires which may occur will
be compelled to announce their own advent and set in motion
the means for their own restraint or annihilation. This re-
quires automatic devices fabricated, installed and finally super-
vised by human beings, and therefore somewhat less certain
to function properly when called upon than is the hand of nature
which produces fires from the requisite combinations of cir-
cumstances. But the difficulties encountered in controlling the
average fire with manually-operated apparatus, even though the
latter may be more powerful because capable of a greater de-
gree of concentration than any automatic equipment which it
is commercially practicable to install, increase rapidly during the
early stages of the fire development, multiplying in much more
than direct relation to the age of the fire ; and on the whole the
automatic equipments, if perfected so far as has been proven
feasible at the present time, are vastly more eft'ective than manual
apparatus in preventing heavy fire losses, even though failure
to control may result in occasional individual instances from
any one of the several causes which may interfere with opera-
tion of an automatic system. Their superiority lies in their
ability to attack each fire in its incipiency, during the period
♦Class of 1900. Professor of Fire Protection Engineeering, Armour In-
stitute of Technology.
THE AUTOMATIC CONTROL OF FIRE.
BY FITZHUGH TAYLOR.*
The problems which confront the fire protection engineer in
the practice of his profession may be classified into two general
groups, more or less allied in certain individual instances, and yet
as a rule quite distinct one from the other. The first comprises
those relative to prevention of fire, and involves observance
of proper methods of construction of buildings, avoidance or re-
moval of unnecessary fire hazards and intelligent segregation and
safeguarding of hazards which are, necessary to the conduct of
certain industries or inseparable from them. The second includes
problems bearing upon control and extinction of fires, and covers
design, installation and maintenance of apparatus and equipment
employed to extinguish fire or oppose its progress.
In dealing with problems of the second group it is in cer-
tain instances possible to provide only manually-operated pro-
tective equipment to be handled during fires by employes of
property owners, municipal firemen or both, but in many cases
it is now practicable to take such measures as will furnish a
reasonable degree of assurance that fires which may occur will
be compelled to announce their own advent and set in rnotion
the means for their own restraint or annihilation. This re-
quires automatic devices fabricated, installed and finally super-
vised by human beings, and therefore somewhat less certain
to function properly when called upon than is the hand of nature
which produces fires from the requisite combinations of cir-
cumstances. But the difficulties encountered in controlling the
average fire with manually-operated apparatus, even though the
latter may be more powerful because capable of a greater de-
gree of concentration than any automatic equipment which it
is commercially practicable to install, increase rapidly during the
early stages of the fire development, multiplying in much more
than direct relation to the age of the fire ; and on the whole the
automatic equipments, if perfected so far as has been proven
feasible at the present time, are vastly more effective than manual
apparatus in preventing heavy fire losses, even though failure
to control may result in occasional individual instances from
any one of the several causes which may interfere with opera-
tion of an automatic system. Their superiority lies in their
ability to attack each fire in its incipiency, during the period
♦Class of 1900. Professor of Fire Protection Bngineeering, Armour In-
stitute of Technology.
The Armour Engineer,
IV— 2. May, 1912.
The Automatic Control of Fire,
Fitzhugh Taylor.
May, 1912] TAYLOR: CONTROL OF FIRE 229
which otherwise would in most instances be consumed in dis-
covering, locating and announcing the fire and in transporting and
placing the apparattis necessary for its control.
The most effective agent at present available for the auto-
matic control of fire is commonly known as the automatic
sprinkler equipment, and it is the purpose of the present paper
to describe in a general way certain representative forms of
apparatus employed in these equipments, and to present some
experimental data of a character not commonly included in articles
on the subject.
A sectional view in perspective. Fig. 1, bound herewith as a
folder, shows a typical factory or warehouse in which is in-
stalled an automatic sprinkler equipment, and is adapted with
a few alterations and additions from a diagram published by the
Factory Insurance Association of Hartford, Connecticut. The
equipment consists of one or more vertical feed pipes called
risers, connected near each ceiling to horizontal feed pipes
known as cross mains, the latter feeding branches equipped with
automatic sprinklers and placed within a few inches of each
ceiling. In the view referred to the riser is shown near the left
row of columns, the crossmains are supported immediately be-
low the girders at each ceiling, running at right angles with the
girders, and the sprinkler laterals or branches are run parallel
to the girders in the center of each ceiling bay. The sprinklers
are essentially nozzles made of composition metal, normally held
pressure tight by parts retained by an especially-compounded.
low-fusing solder which melts and releases the parts under the
influence of undue heat and permits the sprinklers to distribute
water over areas restricted by a deflector or distributor mounted
over each orifice. These devices will be illustrated and described
later in greater detail.
The riser is served by an underground feed pipe to which
at least two reliable independent water supplies should be con-
nected. Fig. 1 shows a wooden tank, usually termed a grav-
ity tank, elevated in a tower above the highest sprinklers as
one of the water supplies. It also shows an underground pipe
leading from a private stationary fire pump. At the right of
the view is shown a third supply pipe not infrequently used in
cities which have public fire departments, known as the steamer
connection. It is carried through the building wall above the
ground and on a side abutting upon a street, is fitted with hose
connections, and is utilized by the fire department by coupling
to it a line of fire hose from the first or second steam fire en-
gine which arrives in response to an alarm. The room im-
mediately below the gravity tank in the tower or elevated tank
230 THE ARMOUR ENGINEER [Vol. 4, No. 2
house frequently contains one or more cylindrical steel pres-
sure tanks of from three thousand to five thousand gallons
capacity, connected to the tank feed pipe and normally kept two-
thirds full of water, air being maintained at from seventy-five to
ninety pounds pressure in the space above the water. A fifth
type of supply which is sometimes chosen by the designing en-
gineer is a connection of liberal size to an adjacent city water
main, provided that the pressure normally maintained on the
main is adequate for sprinkler service.
Each of the independent water supplies is connected to
the underground or riser through a separate swing check valve,
as well as through a gate valve which is normally strapped and
padlocked in wide open position. This arrangement of check
valves permits the riser pressure to be determined always by the
water supply which furnishes the highest pressure. Where pres-
sure tanks are installed in towers or elevated tank houses they
usually impress upon the riser a pressure higher than that of
any other water supplies, but are prevented from delivering
water into the gravity tanks or other supply connections by the
check valves placed in those connections. In case of fire the
sprinklers which open are then at first supplied by the pres-
sure tanks. As the latter discharge their contents the tank pres-
sure is reduced by the expansion of the air which is expelling
the water, and when this reduction has progressed sufficiently
the gravity tank or city service checks open and permit these
supplies to aid in serving the sprinklers. Meantime, if the
private fire pump is placed in operation, or if a steamer couples
to the steamer connection, it is probable that the pump will
furnish water under a pressure sufficient to close all of the tank
and other checks and the sprinklers will thereafter be served
by the pump supply so long as it is available.
The tanks used for sprinkler water supplies demand no
especially detailed description in an article of this character, al-
though the engineer who supervises the installation of a
sprinkler equipment must ascertain that due care has been ex-
ercised in their design, in the selection of materials used in their
construction, in the location and support of the tanks at the time
of erection, in the installation of water level indicators, tank
connections, ladders for inspection and heaters to maintain the
temperature of the liquid contents above the freezing point in
cold weather, and in provision of all practicable safeguards
against deterioration.
Private fire pumps, on the contrary, are especially designed
for fire service and differ in some respects from pumps com-
monly used for other purposes. The type of pump most widely
May, 1912]
TAYLOR: CONTROL OF FIRE
231
used at present for private fire service and preferred by most
fire protection engineers for locations where its use is prac-
ticable is known as the National Standard Steam Fire Pump, and
is illustrated in Fig. 2. Pumps of this type are built by sev-
eral pump manufacturers, and employ no unusual principles of
operation, their distinctive characteristics being mainly details
of design.
The standard fire pump may be described as a duplex,
double-acting, inside-plunger pump, with steam slide valves op-
National Staurtarrt Steain Fire Pump.
erated mechanically from the piston rods through rock shafts
and rocker arms. The steam admission and exhaust passages
are exceptionally liberal in area, as are the aggregate suction
and discharge valve areas in the water end. The pumps are
built in four sizes for discharge capacities of 500, 750, 1,000 and
1,500 gallons per minute respectively, at speeds of one hundred
and twenty to one hundred and forty strokes per minute, or, as
commonly expressed, sixty to seventy revolutions per minute al-
though the designs employ no crank shafts or other revolving
232 THE ARMOUR ENGINEER [Vol. 4, No. 2
parts. The steam cylinders are of cast iron, enclosing cast iron
pistons fitted with cast iron rings, and are equipped with manu-
ally-operated cushion valves for regulating the length of stroke
and amount of steam cushion at each end of the cylinders. Two
of these valves appear in the illustration, immediately below
the steam chest. Each controls a small steam passage between
the steam clearance space and the exhaust passage, thus af-
fording a means of regulating the compression in the clearance
space after closure of the exhaust port by the slide valve. The
slide valve rods are of Tobin bronze, and this material and the
positive rock shaft drive are employed as a safeguard against
stiffness of action after considerable periods of idleness. The
rock shafts turn in bronze bushings for a similar purpose.
In the water end bronze water plungers without packing
are usually employed, although in localities where the water
contains much abrasive material, as is true of water from the
Ohio, Mississippi and Missouri rivers, packed pistons running
in bronze bushings are sometimes used. The piston rods are of
Tobin bronze. Here again the aim is to so construct the pump
that it may stand in idleness for considerable periods and yet
be in readiness to deliver its full capacity on short notice. The
water valves are of bronze, with rubber discs, and are of the
poppet type, lifted by impact of the water and closed by brass
springs. They are grouped in multiple on the suction and de-
livery decks in a ratio of about three to two. The aggregate valve
area is liberal in proportion to the plunger area to yield full
capacity at comparatively high speeds. The water end of the
pump is fitted with unusually large air chambers on the suc-
tion pipe and delivery chamber to minimize pulsation in the
discharge pipe and water hamm.er in the suction chamber at
high speeds. Gate valves, for attachment of two and one-half
inch hose are also attached to the discharge chamber.
The ratio of steam to water areas in standard fire pumps
varies from about four to one in the smallest size to about two
and three-fourths to one in the largest, in order that good fire
pressure may be developed by boiler pressures not exceeding
fifty pounds per square inch. The entire construction is ex-
tremely rugged and heavy, to withstand on the one hand sud-
den admission of steam without preliminary warming of the
castings and on the other hand the shocks incident to operation
under water pressures of two hundred to two hundred and
twenty-five pounds per square inch, which may be obtained from
these pumps where steam at from eighty to one hundred pounds
pressure is available, and which are sometimes desired at fires
where water must be delivered through long lines of fire hose.
May, 1912]
TAYLOR: CONTROL OF FIRE
233
Another type of pump which has been used to a consid-
erable extent for sprinkler supplies in water-power mills is that
known as the rotary or gear pump. A pump of this type espe-
cially designed and fitted for fire service is illustrated in Fig. 3.
Its principle of operation is the same as that of the small gear
Fig. 3. National Standard Rotary Fire Pump.
pumps now quite widely used for circulation of cooling water
through the cylinder jackets of internal-combustion engines, and
involves the use of two parallel shafts geared together and mov-
ing in opposite directions, each carrying within the casing a
large cam or pinion, these meshing one with the other, trap-
234
THE ARMOUR ENGINEER
[Vol. 4, No. 2
ping water between the gear teeth and the wall of the casing
and delivering it upward into the discharge chamber. Fig. 4 is
a sectional diagram of a pump of this type. Rotary pumps have
been utilized for fire service mainly in properties equipped
with water power and lacking steam plants suitable for driv-
ing steam fire pumps. They can be readily driven by water
wheels or from countershafts and have been reasonably suc-
cessful in these locations.
The tendency of urban property owners to adopt central
station electric power has created during recent years an active
demand for fire pumps suitable for electric driving without
speed reduction, and centrifugal pumps of the multistage patterns
are now being perfected for this service. It is in certain
instances possible to secure a fairly reliable pump supply for a
Fig. 4. Typical Diagramiiintic Section of Rotary Pire Pump Cylinders,
sprinkler equipment by utilizing an electrically-driven centri-
fugal pump where the available power facilities would not per-
mit the use of a pump of the positive-displacement type, but
pumps of the latter class are more flexible in their performance
and are therefore preferred where it is found practicable to
make use of them.
Perhaps the simplest imaginable expedient for compelling
a system of piping, normally under water pressure and ar-
ranged as has been illustrated and described, to discharge water
automatically upon any fire which might subject it to abnormal
temperatures would be to perforate the piping and fill the per-
forations with some fusible material, such as solder, which when
cold would have sufficient strength to resist the water pressure
and remain tightly in the openings in the pipe. But fusible plugs
May, 1912]
TAYLOR: CONTROL OF FIRE
235
of such a nature would necessarily be directly in contact with
the liquid contents of the piping, and under the influence of fire
could be heated very little if any more rapidly than the water
and the piping. The resulting delay in delivery of water would
in many instances be fatal to control of the fire, and to minimize
this delay the outlets of sprinkler piping are equipped with
devices which will release the water when an abnormal rise of
temperature occurs, but which are designed in such a way as
to remove the fusible elements from the cooling influences of
the water and piping.
Automatic sprinklers, or sprinkler heads as they are col-
loquially termed, are of various forms, but usually of one of
two general types. Fig. 5 is a sectional view of one of these,
and sprinklers of this general type, diflfering considerably how-
Fig. 5. Section of Typical I^ever Type Automatic Sprinkler.
ever in design, are now manufactured and installed by several
sprinkler equipment companies. The body, B, of this sprinkler
is a casting of composition metal. The lower end carrie? a
one-half-inch pipe thread and is cylindrically bored to form
a vertical nozzle. Extending upward from diametrically opposite
points on this vertical tube are two arms, A A, which join over-
head and are drilled and tapped at the junction to receive a
vertical retaining and adjusting screw, C. The nozzle orifice is
normally closed by a disc or cap, D, composed of several pieces
of metal so formed and joined as to exhibit elasticity to a
marked degree. The cap is held upon the nozzle by a pair of
toggle levers, T, which are forced downward upon the cap by
the adjusting screw in the top of the frame, and whose outer
ends are held together by a fusible link, E, consisting of two
236
THE ARAIOUR ENGINEER
[Vol. 4, No. 2
or more plates of composition metal soldered together with a
specially compounded solder. A distributor or deflector, F, is se-
cured to the top of the frame by the adjusting screw.
The drawing shows that the bearing of the toggle levers
upon each other does not lie in the center line which passes
through their bearings upon the cap and the adjusting screw.
When the latter is set downward, therefore, to press the toggle
system against the cap, the soldered link is stressed in tension
by the tendency of the toggle levers to separate at that point.
The vertical arms of the frame are also stressed by the re-
action from the adjusting screw, and this stress, as well as that
in the link, is present in every sprinkler of this type as it stands
in service. The effect of fire upon the head is the fusing of the
Section of Spriukler
Shown hi Fig. 7.
A Lever Type Automatic
Sprinkler.
solder in the link, whereupon the stresses resident in the frame
will usually scatter the toggle levers and parts of the link for
a distance of several feet, permitting the cap to be blown from
the nozzle by the water pressure. The jet which issues from the
nozzle strikes the upper portion of the frame and the deflector,
and is thereby broken up and distributed.
Fig. 7 shows a sprinkler of the type just described, dif-
fering in details of design from that of Fig. 5. Fig. 6 shows the
same head with the working parts removed and the lower por-
tion of the frame sawed in section to show the form of the
nozzle waterway.
Another sprinkler which is widely used is illustrated in
Figs. 8 and 9. One view is a section. The body, B, screws into
May, 1912] TAYLOR: CONTROL OF FIRE 237
the tee in the Hne pipe; the yoke or frame, A, screws into the
body and carries the deflector, J. Between A and B is held a
flexible metal diaphragm, C, with a half-inch hole in the cen-
ter. Into this hole fits the hemispherical glass valve, E, held in
position by a small metal cap and a strut of three pieces, F, G rind
H. These three parts are held together by soft solder fusing
at about 166 degrees Fahrenheit.
When the temperature of the solder reaches this point the
disruption of the strut begins, and ultimately takes the form of
a rocking motion one part about the other. During this move-
ment the flexible diaphragm with the water pressure under its
entire area holds tightly against the glass valve until as the
strut finally falls apart both valve and strut are thrown out
Fig. S. A Diaphragm Type Automatic Spriukler.
into the room ; then a stream of water, striking the deflector, is
scattered in all directions.
The solders most commonly used in the manufacture of
sprinklers have fusing temperatures of about 165 degrees Fahr-
enheit, and these sprinklers are known as standard degree heads
and are used for all ordinary locations.
Certain locations such as boiler rooms, dry rooms- and
spaces under skylights which are frequently warmed by rise of
warm air from below and by the direct rays of the sun from
above demand heads made with less sensitive solders, and sev-
eral grades of harder solder are regularly used for the purpose.
While a standard degree sprinkler, when slowly heated in
water or oil, will usually fuse when the temperature of the
238
THE ARMOUR ENGINEER
[Vol. 4, No. 2
submerging liquid reaches approximately 165 degrees it must not
be supposed that in the average fire the heads will fuse im-
mediately upon the attainment of that temperature by the sur-
rounding air, because the specific heat of air is much less than
that of the liquids mentioned, and air is correspondingly slower
in conveying to the soldered links the quantity of heat neces-
sary to raise their temperature to the fusing point.
Several methods may be used to demonstrate the sensi-
tiveness of an automatic sprinkler. For example, a standard
degree sprinkler of the type shown in Fig. 5 may be held in
one hand and a parlor match in the other. If the match be
ignited, and held at such an angle as to produce the maximum
duration of flame and keep the soldered link continuously en-
Fig:. fl. Section of the Sprinkler Slio^vn in Fig, 8.
veloped in the flame, the sprinkler may possibly open before
the flame ceases, but usually unless the experimenter .has ac-
quired some skill by practice a second match is required to fuse
the link. In this simple test the link is plunged into a small
flame whose temperature is perhaps 2,000 degrees Fahrenheit,
and is apparently slow to operate. On the other hand, a similar
sprinkler dipped bodily into a kettle of water or oil whose tem-
perature is 175 degrees Fahrenheit will open in from five to
ten seconds. If one or two liberal armfuls of dry excelsior be
piled and ignited on the floor of a room equipped with sprinklers
it is probable that within a few seconds the flame will reach a
sprinkler overhead, and a standard degree head enveloped in a
flame of such a volume may readily open in from ten to twenty
seconds after the first flame contact. In contrast to this ac-
May, 1912]
TAYLOR: CONTROL OF FIRE
239
tion, a test fire built of comparatively coarse material with only
sufficient kindling to insure its ultimate growth will gain head-
way slowly without discharging large quantities of flaming
volatile matter, and may burn for several minutes before fusing
the link of a sprinkler directly overhead.
The rate at which an automatic sprinkler discharges water
after being opened by fusing of the solder is obviously de-
pendent upon the water pressure to which it is subjected. Fig.
10 displays a curve which represents the relation between water
so
1
^
^
30
^
J
y
^
5
ZO
^
y
1
/
/
«/
/
/
/
/
o
S-
0
3
o
4
0
S
o
eo
Water Pressure in Lbs. per a In.
Automatic Sprinkler o£
Fig. 10. Tjiiical Discharge Characteristic of a
tlie Diaplirag-ni Type.
pressure and rates of delivery obtained experimentally from a
sprinkler of the pattern illustrated in Fig. 8, and this curve is
fairly representative of the performance of most modern auto-
matic sprinklers. Sprinklers may very fairly be considered ef-
fective as fire extinguishing agents when discharging water un-
der a pressure of fifteen pounds per square inch at the nozzles,
if spaced over the ceiling in accordance with standard practice,
but somewhat higher pressures are preferred and are available
in a majority of sprinklered buildings. Many fires have doubt-
less been controlled or extinguished by sprinklers on upper
240 THE ARMOUR ENGINEER [Vol. 4, No.
floors of buildings served only by gravity tanks, where the
available pressures were considerably less than that quoted.
Fig. 11 illustrates an automatic sprinkler in operation, the
view showing the device mounted upon a test stand especially
arranged for study of the water distribution and provided with
reference circles ten feet in diameter, mounted on vertical rods
as aids to the observer in judging the floor and ceiling areas
covered. Most sprinklers yield a somewhat ' more scattered dis-
tribution with a lower central density than that illustrated, which
1
Fig. 11. Automatic Sprinkler in Operation on Test Stand.
is shown discharging at a pressure of about forty pounds per
square inch.
Sprinklers installed in the lower stories of sprinklered
buildings are necessarily subjected to heavier water pressures
than those in the upper stories, because of the greater gravity
heads impressed upon them by the column of water in the feed
pipes or risers. The fire hazards, however, are not necessarily
more severe in the lower stories, and in fact the reverse may
be true in many instances, and if sufficient pressure is pro-
vided to produce effective discharges at the highest line of
sprinklers these rates of discharge must be considered adequate
as well for any of the lower floors, and the excess • deliveries
caused by the heavier pressures in the lower stories may be
May, 1912]
TAYLOR: CONTROL OF FIRE
241
fairly counted as wasteful of water. This extravagance in the
use of water on the lower floors is not considered of sufficient
importance to warrant the complication of apparatus which
would be involved in the use of automatic pressure-reducing
valves or of sprinklers of varying orifice diameters, but a thesis
recently completed by students of the Department of Fire Pro-
tection Engineering at Armour Institute of Technology suggests
an interesting possibility which may be worthy of investigation
over a range of conditions wider than that already covered.
w ■
^
•I
^
^
^-v^
=#=
^
>
e^
^^
1
J
y
<r
^
•s
V
/
1
/
t
^
/o
g
/
p .,
s,
0
3
o
4
o
^
o
^0
Water Pressure in Lbs. per a in.
Fig. 13. Discharg-e Characteristit's of Sprinklers Differing in Lengtli of
Slioulder at Orifice.
The thesis in question included as one of its divisions the
determination of discharge characteristics of a group of sprinklers
of the pattern shown in Fig. 6, these sprinklers having been'
made especially for the tests. They dififered from ordinary
sprinklers arid from each other in the lengths of the shoulder
at the orifice, which is clearly apparent in Fig. 6. The shoulder
length varied in the different members of the group from slightly
more than that shown in Fig. 6 to about one-half the length
of the inlet waterway. The discharge characteristics of these
specially-formed sprinklers are shown in Fig. 12. The explana-
242 THE ARMOUR ENGINEER [Vol. 4, No. 2
tion advanced for the aberration of the curves was that at the
lower jet velocities the efflux coefficients were influenced by the
diameter of the waterway down stream from the shoulder, and
at the higher velocities principally by the contraction from the
shoulder. It was found that the longer shoulders required higher
velocities than the shorter to produce the aberration.
The results of these tests have suggested the thought, not
as yet confirmed by experiment, that sprinkler nozzles designed
with two or more shoulders in the waterway might be found
to show corresponding points of aberration under continuously
increasing pressures, and that such sprinklers, showing several
depressions in the discharge characteristic, would be less ex-
travagant than those at present used in their consumption of
water in the lower stories of high buildings, and would at the
same time yield in the upper stories the efflux coefficients which
are now accepted as standard. This theory is advanced with some
diffidence, and with the thought that it may provoke further
experiment.
The effectiveness or protective value of an automatic sprink-
ler equipment is dependent to a very considerable degree upon
the location of the sprinklers with respect to each other and to
the ceilings. In order to be effective as fire extinguishing de-
vices sprinklers must be capable of opening promptly in the
presence of fire, and of wetting adequately the ceiling and floor
areas allotted to them by the designer.
In designing a sprinkler equipment care is first taken to
devise an arrangement which will be favorable to prompt open-
ing of the heads in case of fire. To attain this end all prac-
ticable expedients are employed to compel the products of com-
bustion to bank or accumulate at the ceiling directly over their
source. Ceilings such as those shown in Fig. 1 promote prompt-
ness of sprinkler operation because the girders divide them into
channels or bays, any of which may act to retain a thick stra-
tum of hot gases from a fire under it, a condition favorable to
operation of sprinklers in that bay. Hot gases may fill one bay
and underflow the girders, thus attempting distribution over the
ceiling transversely to the girders ; but each successive bay as it is
entered by the gases tends to retard their further progress until
it is filled, with the result that the gases are to a considerable ex-
tent retained by the ceiling directly over their point of origin,
which is the point demanding sprinkler service. Windows which
extend to the ceiling line, and open elevator and stair wells and
hatches, are causes for apprehension, because they afford the hot
gases means of escape from the ceiling localities where it is de-
sirable that the gases should be retained. When sprinklers are to
May, 1912] TAYLOR: CONTROL OF FIRE 243
be installed in buildings whose ceilings afford escape openings
of the character mentioned, it is wise to build aprons or curtains
around the openings if the latter can not be entirely enclosed,
to aid the floor beams in retaining hot gases. Obviously, scant
clearance between ceilings and sprinklers makes for promptness
of operation, and the heads are therefore mounted as close
to the ceiling as is permitted by the character of the ceiling and
the requirement of good water distribution.
After adequate provision has been made for banking of heat
around sprinkler laterals the designer must decide upon distri-
bution and arrangement of the heads. It is common practice to
assign from eighty to one hundred square feet of floor or ceiling
area to each sprinkler. The best arrangement is that shown in
Fig. 1, where at least one lateral is installed in each bay, lying
parallel to the girders; but the cost of this arrangement is
prohibitive for joisted ceilings having joist channels only one or
two feet in width, and in such cases the laterals must run trans-
versely to the joists. Care must then be observed in location of
the heads on the various laterals, for unless every joist channel
can receive water from one or another of the laterals it is pos-
sible for fire to travel unchecked along channels which are not
so protected.
Buildings whose occupancy is such as to require that they
be heated during all seasons are equipped with what are known
as wet pipe sprinkler systems, the distributing piping being main-
tained under full water pressure at all times. In buildings such
as warehouses which are not well heated during cold weather
such systems would suffer from freezing of the contained water,
and buildings of this class are therefore equipped with dry pipe
systems. The arrangernent of sprinklers and piping is very simi-
lar in the two types of system, although in piping a dry system
the pipes are not infrequently pitched toward the riser to a
slightly greater degree than in wet systems to facilitate drainage,
and the sprinklers are invariably installed in the upright posi-
tion to avoid retention of water in their nozzles, a practice which
is sometimes departed from in installation of wet systems al-
though it is preferred for the reason that most sprinklers yield
a somewhat better water distribution in the upright position
than when pendant.
The dry system differs from the wet principally in that at
the base of the riser is installed an automatic valve which is
subjected upon its lower side to water pressure from the un-
derground supply pipe and on its upper side to a considerably
lower air pressure which is normally maintained throughout the
distributing system of pipes. These valves are known as dry pipe
244 THE ARMOUR ENGINEER [Vol. 4. No. 2
valves, and possess differential properties by virtue of which
they may be held closed against the water pressure by
a considerably lower air pressure in the riser, and are
so constructed as to open automatically and admit water to
the risers which they control when the compressed air is
released from the sprinkler piping by opening of one or more
heads. Most buildings have at least one or two rooms which are
heated in cold weather, and in which dry pipe valves may be in-
stalled without danger of freezing of the water in the supply
pipe; in buildings which are unhealed throughout, small wooden
closets or brick vaults are built around the dry pipe valves, and
in cold weather a lantern or small heater in each enclosure pre-
vents freezing within the supply pipe.
Sprinklers on wet systems deliver water more promptly
after fusing than on dry systems, the air contents of systems of
the latter type requiring an appreciably longer time for expul-
sion before water reaches the open heads. For this reason wet
systems are selected for all locations where danger of freezing
is not present, and in some localities it is accepted practice to
maintain dry systems only during the cold seasons, these equip-
ments being maintained as wet pipe systems during the remain-
ing months of the year.
Dry pipe valves may readily be fitted with attachments
which will cause fire alarms to be sounded automatically by the
action of the valve mechanism at the time of opening, either by
actuating an electric circuit-closer or circuit opener, or by ad-
mitting water to a small Pelton wheel which may drive the
tapper of a gong. These valve attachments are at the present
time fairly reliable as means of producing automatic fire alarms,
provided the signaling apparatus is maintained in operative con-
dition. Automatic alarm valves are also made for installation in
risers of wet pipe systems, generally utilizing the action of a
specially-designed swing check valve to actuate the fire alarm
apparatus. They are widely used, and are fairly successful un-
der favorable conditions, but they are less reliable than most of
the other apparatus employed in sprinkler equipments at the
present time.
Automatic sprinkler equipments, contrary to a common un-
derstanding of their functions, are not relied upon primarily by
the fire protection engineer to carry the entire burden of ex-
tinction of fire. Many fires, especially those resulting from the
phenomena popularly grouped under the classification of spon-
taneous ignition, originate at points which are not readily
reached by the discharge of sprinklers, although every effort is
made by the designing engineer to limit these inaccessible places
as far as is possible. Some fires, such for instance as those which
May, 1912] TAYLOR: CONTROL OF FIRE 245
are caused by slow heating within a pile of neglected rubbish,
demand for their complete extinction that the pile of fuel be
forcibly scattered and deluged by a powerful stream, and sprink-
lers are obviously incapable of furnishing service of this char-
acter. The task allotted to the sprinkler equipment in every such
instance is to discover the location of the fire, signal for aid
through the medium of the fire alarm attachments previously
mentioned, and by distributing water over and around the fire
hold the latter in restraint until assistance arrives. If the equip-
ment successfully fulfills these of its obligations it is entitled
to full credit. Very frequently, however, fires are completely ex-
tinguished by sprinkler equipments within a few minutes of their
origin.
The fire record of automatic sprinkler equipments is a re-
markably creditable one. For a number of years the National
Fire Protection Association compiled and published in the records
of its annual conventions tabulations based upon reports by its
members and others covering fires in sprinklered buildipgs
throughout the country. The reports covered equipments known
to be faulty in arrangement or character of apparatus, as well
as those considered worthy of thorough confidence, and in-
cluded records of fires in many classes of property ; and not long
ago, when the total number of fires reported exceeded six thou-
sand, the failures to control by sprinklers aggregated only be-
tween six and seven per cent of the total reported fires which
had been successfully controlled by sprinkler equipments. Of the
ninety-three per cent reported as successful about one-third
called for the service of only one sprinkler each, and more than
one-half involved three sprinklers or less. Careful analysis of
the six per cent of failures shows that a large proportion of this
number yielded explanatory evidence of the causes of the
failures, such for instance as closed gate valves in the water
supply pipes or recognized defects in the equipments, and the
obvious deduction from such an analysis is that there are very
few fires which can not be held within bounds by correctly de-
signed and properly maintained automatic sprinkler equip-
ments.
PURCHASE OF COAL ON SPECIFICATIONS.
BY W. O. COLLINS.*
During recent years the purchase of coal along scientific
lines has received a great deal of consideration, but only of late
have demonstrations been made which would justify any one
in saying that the B. T. U. system was a real success and
certain to succeed.
Even now the system has many obstacles to contend with in
the form of opposition from the coal trade and skepticism due
to some inefBcient methods used in sampling and testing and
the great human element of uncertainty in the testing engineers.
Under the B. T. U. specifications the purchaser in reality
buys heat units rather than tons of coal, although the coal is
actually paid for by the ton, but the price is determined from a
calculation based on the chemical analysis of coal delivered.
Considering the wide use of coal, the vast amount of money
involved in the industry and the many and wide variations in the
quality and characteristics of coal, it seems strange, when we
stop to consider it, that the purchase of this material along mod-
ern lines has been slow in starting and developing.
Up until a few years ago the larger consumers employed the
boiler test to determine whether or not coal was efficient or up
to contract requirements. Selections of coal for contracts were
frequently made by this method. Coal contractors were re-
quested to make a shipment of coal representative of the fuel
which they proposed to furnish if awarded the contract. Sev-
eral shipments so received were subjected to burning tests under
the boiler and evaporation per pound of coal and the cost to
evaporate 1,000 pounds of water was determined with greater
or less accuracy depending on the care with which the tests were
made.
On public and political contracts the evaporation method
has caused no end of criticism, as there are many conditions
under the control of the testing engineer and fireman by means
of which the results can be controlled at will. Furthermore,
even if the tests are honestly and efficiently made, they are
useless in the case of a legal fight, as it is always possible to
show that the conditions of testing are constantly changing to
a greater or less extent, due to the formation of boiler scale,
weather, load and firing requirements.
♦Class of 1902. Vice-President, Gulick-Henderson Company, Chicago.
May, 1912] COLLINS: PURCHASE OF COAL ^ 247
Alon^ with and following this method of specifying and
regulating deliveries a chemical analysis showing the amount of
moisture, volatile matter, fixed carbon, ash and heat value was
frequently incorporated in the contract together with the guar-
antee of evaporation obtained by the boiler test method. This
was often a strengthening clause and was many times the
basis of making settlement where substitution was clearly
evident.
It cannot be said that any of these methods were ever uni-
versal to any extent nor is the new and improved B. T. U. sys-
tem in universal use, for in many cases the fuel which forms
forms from ten to twenty-five per cent of the yearly expense, is
bought without any supervision whatever, while the much less
expensive items, such as steel, pig iron, cement, electrical mate-
rials, paper, etc., are often purchased on the most rigid speci-
fications and guarantees.
Following the public demand for efficiency and honest pur-
chasing the political and public institutions have in many cases
been the leaders in scientific methods of purchasing coal. Thus,
in 1907, the United States government adopted a form of
B. T. U. specifications which is now in use by practically all
government departments. The methods used by the govern-
ment and the methods now in use by other consumers are.
generally, based on the same fundamental principle, which is the
"delivery of heat units." While there are several methods of
regulating and figuring the value of a delivery, practically all
of them consider the analysis of as much importance as the
weight of the coal.
Our concern is now setting the price on $1,500,000 worth
of coal delivered annually to at least 400 consuming plants,
among them all of the Cook County institutions, stations of the
Sanitary District of Chicago, power and heating plants of the
South Park Commissioners and the West Chicago Park Com-
missioners, and all schools under the Board of Education, as well
as many private plants, loop buildings and power plants in
smaller cities.
Our recent work for the Merriam Commission investigat-
ing the various purchases in the City of Chicago, was beneficial
in showing the abuses which often gradually grow out of the
old or slack methods of purchasing.
The result was that the men in actual charge of these
purchases welcomed an opportunity to purchase their coal supply
on a specification and under a system which would not only give
them good coal at a low figure, but also relieve them of the re-
sponsibility or the criticism due to the continued and expensive
248 THE ARMOUR ENGINEER [Vol. 4, No. 2
abuses under the old and more common system. Since the
adoption of the Merriam Commission recommendation in this
respect the PubHc Works Department of the City of Chicago is
getting a better grade of coal and saving hundreds of thousands
of dollars by the elimination of unbusinesslike methods, so that
today nothing but commendation is heard from even the coal
trade itself in connection with this big item of expenditure.
The specifications as we prepare them differ in detail for
different institutions, due to the variations in the coal require-
ments and business methods of the office. All embrace clauses
to cover points relating to grade of coal, point and time of
deHvery and other special requirements, and it will be evident
that a specification should cover something more than the mere
physical properties of coal.
After the bids are received they are tabulated. The bidder
who guarantees the greatest number of heat units for one cent
is the lowest bidder.
After the contract is let, deliveries are sampled at frequent
intervals and analyses run on combined samples and from these
analyses the delivered value of the coal is calculated in accord-
ance with the terms of the specifications and contract.
The specifications state the high and low limits of analysis
which will be accepted under any conditions. Coal accepted is
paid for on the showing of the analysis.
The method of sampling, chemical analysis and other de-
tails of the process are now fairly well standardized and while
there are still differences of opinions in minor details, neverthe-
less it is a fact that they are as well standardized and can be as
accurately handled as in the sampling and testing of other
materials of commerce, such as iron, steel, cement, etc.
The B. T. U. system has many advantages, especially for
public bodies and large purchasers where outside influences
are liable to interfere.
First. Bidders are all placed on exactly the same basis for
consideration. Since awards should be made on the basis of the
maximum number of heat units for one cent, there can be no
possible controversy if this rule is followed. For example, bids
are received from three different bidders on Illinois or Indiana
lump. No. 1 bidder agrees to furnish "Atlas Lump" on the
basis of 75,000 B. T. U. for one cent. No. 2 agrees to furnish
"Perfection" lump with 100,000 B. T. U. for one cent, while
No. 3 agrees to furnish "Economy Lump" with a guarantee of
125,000 B. T. U. for one cent. At a glance a child could tell
which is the cheapest and best bid. Under the old method the
bidder offering "Perfection Lump" would have a big advantage.
May, 1912] COLLINS: PURCHASE OF COAL 249
Second. Since only price and quality enter into the calcu-
lations upn which awards are made it will be evident that
"trade names" have no influence whatever. Thus it is often pos-
sible for dealers to offer coal of good quality from small and
comparatively unknown mines and if the bidder is responsible
such bids can be accepted without any possible chance of loss.
Third. The consumer is insured against the delivery of
poor coal, since the penalties tend to stimulate the delivery of
only the best coal to those plants where regular tests are made.
Fourth. A specific and equitable basis of payment is pro-
vided for should coal be below grade. Coal rejected under the
old methods of contracting was often accepted at contract price
and burned up simply because of the delay of getting it removed.
Fifth. A definite basis for cancellation of contract and
otherwise regulating of deliveries, etc., is provided.
Sixth. Constant testing and inspection has a healthful
influence on the plant. By the means of the results operating
engineers are enabled to get better efficiency from their men.
Furthermore, the knowledge that constant and regular testing
is being done stimulates the best efforts of the contractor to
furnish uniform coal.
Seventh. The system can be cheaply and efficiently applied
at a cost well within the limits for inspection and testing and
seldom exceeding one to two cents per ton.
Of course, it may be truly said that the coal trade as a
whole opposes the system. Some of the dealers oppose it be-
cause it eliminates their chances for the substitution and deliv-
ery of poor coal. Other operators producing low coal naturally
are at a disadvantage and also use their powerful influences to
return to the old hit-and-miss methods. Few of them stop
to consider how it has really opened up a field for fair and
equal competition especially on public business. In general,
however, all this opposition is and has been a boost, for it is
plain to be seen that their opposition is based on reasons of per-
sonal gain and if they really favored the system too strongly
the purchasers would not want it.
With all of the opposition from this source, however,
there are always plenty of bidders. Generally there are more
than were received before when bids were taken under the
old methods.
The Board of Education of Chicago formerly received only
three or four bids and some of these were from affiliated com-
panies. Last year there were twenty-nine (29) bidders all
independent and actually after contracts. The South Park
Commissioners received ten bids for a much smaller amount of
250 THE ARMOUR ENGINEER [Vol. 4, No.
coal and the West Chicago Park Commissioners using still less
coal received fourteen bids.
It is often said as an argument against the adoption of the
system, that the prices will be raised to cover the element of
gambling due to possible variations in the coal. This most
certainly has not been the result as far as we have been able to
observe. In fact, the price per ton has usually been lower
for the same grade of coal than when the old-style methods of
purchasing were employed. The increased competition governs
this to a large extent.
The savings by the installation of this system in the Board
of Education of Chicago has been figured to exceed $100,000.
The greater part of this saving is due to the increased compe-
tition and consequently lower prices due to the assurances of
fair treatment afforded by the method. A part is due to the
penalties deducted for the delivery of inferior coal and no ac-
count is taken of the large unknown saving by the use of the
better grade of coal received.
Similarly, great savings are being made in other public and
private plants.
Therefore, judging from the continual growth of our own
part of this work and since we constantly hear of the success
of others and of new large consumers starting to take bids on
this or a similar basis I feel safe in predicting a gradual and
steady growth and improvement in the system.
THE MIETZ AND WEISS OIL ENGINE.
BY E. E. MAHER.*
The statement is often made that America is the most ex-
travagant nation. The great natural resources of the United
States have made it easy for its population to produce more and
consume more than less favored people. The small economies
that are necessary for other peoples to practice to exist have
been, except within the limits of certain industries, unknown.
Our forests were converted into lumber with the idea of
making big money quickly and with very little thought of the day
when it would be essential to conserve the remaining timber and
to "harvest" it with the same regard for its continuous produc-
tivity as that given by a careful farmer to his cornfield.
In the same manner virgin soil of wonderful fertility has
been exhausted by growing the most productive crop season
after season, the idea being that when the old farm was ex-
hausted there was plenty of room out West.
"High grading" has been a common practice in mining for
the precious metals and the spoil piles beside our coal mines
contain millions of tons of valuable fuel.
Fire losses are enormously higher than in any other country.
The American merchant and manufacturer, competing in
the world's markets, has been able to obtain his share of trade,
although paying higher wages than his competitors, because his
materials cost him less. The greater earning capacity of some
American workers has been an important factor. Question —
With the proportion of foreign born "hunkies" employed in our
shops increasing, how long will this be a controlling factor?
We have been working on the same principle that a certain
manufacturer followed. He developed a profitable business
manufacturing machinery. He saw an opportunity to secure a
far larger portion of the business than he had heretofore se-
cured. He obtained the capital, greatly enlarged his plant and
his selling organization, and obtained the larger portion of busi-
ness he had desired. But, although he had been an efficient
head for a small undertaking, he was not able to conduct "big
business." Sales were large and profits small. Rather than con-
fess his inability to direct the business, he resorted to subter-
fuge to conceal the real situation. Repairs to tools and machin-
ery went on the balance sheet as "extensions to plant;" losses
♦Formerly of Class of 1905. Vice-President, B. M. Osbun Company,
Chicago.
252
THE ARMOUR ENGINEER
[Vol. 4, No. 2
on goods sold to "development of new machinery," and to "pat-
terns" and so on. Additional capital stock was sold from time
to time, part of the proceeds being used to repair the ravages of
unsuccessful management on the company's working capital. The
day of reckoning finally came, and the stockholders found that
they had been paying in their money to receive it back in divi-
dends, but that the principal no longer existed.
Horizontal Single Cylinder Oil Engine.
The American manufacturer who, with the advantage of
cheap raw materials, is able to maintain or increase his business,
but who disregards opportunities to decrease his manufacturing
cost, is working on a plan as radically wrong as that in the illus-
tration given. It is just as necessary to keep expense in the
manufacturing department down to the minimum as it is in the
purchasing department, or in the sales force.
With the advantage of location and opportunity, it is not
May, 1912] MAKER: M. & VV. OIL ENGINE 253
remarkable that we have overlooked such things as more effi-
cient prime movers, of which the crude oil engine is perhaps
the best and certainly the least used in this country. It is ^
common occurrence for an industry located near an oil field to
use the most inefficient form of steam engine consuming fifty
or sixty pounds of steam per horsepower hour, with coal at
four dollars or more per ton, or more than one cent per horse-
power, when an equally reliable power plant, consuming fuel
oil would deliver a horsepower for less than a third as much.
In Europe a dozen types of oil engines have been developed
and widely used. Of one type, more than 250,000 horsepower
have been built in the last three years and for both stationary
and marine work they are rapidly demonstrating their superiority.
There have been a number of reasons why American manu-
facturers have been slow to take up the oil engine as a prime
mover. In the first place, steam engines and boilers are com-
pletely standardized. A user requiring new equipment can specify
the horsepower he requires and the general type of machine
and the bidder who makes the best guarantee and the best com-
mercial proposition wins.
Gasoline and gas engines were the first forms of internal
combustion engines. They have, to a certain extent, paved the
way for oil, but have also put a number of difficulties in the way
of the man who exploits oil engines. It has been stated that
over eighty-five per cent of gasoline engine trouble has been on
account of radical defects in manufacture or in failure or bad
adjustment of carbureter or ignition devices. The natural con-
sequence of this, because failures were very numerous, is that
the power user, wanting first of all reliable power, specifies
steam equipment to get away from the troubles which he be-
lieves are inherent in all internal combustion engines.
As far as oil engines are concerned, it is a fact that four-
cycle engines, while having some of them very high efficiency,
are usually very complicated in construction and require an ex-
pert to run them. This has been another of the misfortunes of
the oil engine in this country, as one of the first manufacturers
of oil engines pushed his product, a very complex four-cycle
engine, with much more zeal than discretion, and, as a result,
a number of unsuccessful installations gave the industry a "black
eye."
The De La Vergne Machine Company, in the last two years,
has been manufacturing a four-cycle engine and getting very
good results, but this was only after they had manufactured
the well known Hornsby-Akroyd two-cycle engine for a number
of years.
254
THE ARMOUR ENGINEER
[Vol. 4, No. 2
The user of oil engines has very good reasons to consider
a two-cycle engine as more reliable and better adapted for an
isolated plant than a four-cycle engine. There are a number of
Vertical Single fyliniler Oil Kiigine.
types of two-cycle engines which can be relied upon with a
minimum of care to run continuously without expert attention.
As an example, a Mietz & Weiss oil engine on the Pacific
Coast ran four months without once stopping, at full-load, and
May, 19121 MAKER: M. & W. OIL ENGINE 255
a number of Hornsby-Akroyd engines, as well as Mietz & Weiss
engines have been used by the United States government for
light house and other isolated plant service, where a failure of
power would have disastrous consequences.
In America, three rather distinct types of oil engines have
been developed, and two are now well known through a consid-
erable number of successful installations. Some years ago, Ar-
mour & Company, whose great growth has perhaps been as much
due to their care in selection of efficient machinery for use in
their business as to their selection of capable men to supervise
their business, made an investigation to determine the merits of
oil engines. After considering both foreign and American-made
engines. Armour & Company purchased a Meitz & Weiss engine.
Its satisfactory operation has since induced them to purchase
perhaps a dozen engines of the same make. Mietz & Weiss en-
gines ranging in size from twenty-five to one hundred horsepower
are now in use in the various Armour plants. Mietz & Weiss
engines have been put in places where even a temporary shut-
down means a definite money loss to the company, and their
success is an illustration of the wisdom of Armour & Company
in their business enterprises. The country home of Mr. J. Og-
den Armour at Lake Forest, 111., is equipped with three Mietz
& Weiss oil engines, two of ninety horsepower and one of
twenty-five horsepower, which furnishes electric current for
lighting and power for the estate. There is a two-and-a-half
horsepower Mietz & Weiss engine on the testing floor at the
Armour Institute of Technology.
The following describes the principles of operation of the
Mietz & Weiss machine, which is appHcable with slight varia-
tion to any other American two-cycle engine :
All of these engines have a bulb, or some similar contri-
vance, which is heated for five or ten minutes before the engine
is started by means of a kerosene or alcohol torch. After the
engine is started, the heat of compression and ignition keeps the
bulb hot.
In directions for starting the Mietz and Weiss engines, the
catalogue shows the following:
The Mietz & Weiss ignitor is a hollow cast iron ball at-
tached to the cylinder head by a flange and screws, with a cop-
per covered asbestos gasket to make a tight joint. This ball
has a lip or tongue at its open end, projecting into the cylinder,
beyond the cylinder flange, directly in the path of the oil injec-
tion. The oil, coming from the injection nozzle, strikes the
tongue with sufficient force to spray and vaporize it instantly,
forming, with the air and steam in the cylinder, the explosive
256
THE ARMOUR ENGINEER [Vol. 4, No. 2
charge, as the piston completes its compression stroke. The
ignitor ball must be heated before the engine can be started,
and for this purpose there is furnished with the engine a kero-
sene torch to be used about five or ten minutes, which is gen-
erally sufficient. After the engine is started and the ignitor ball
has a dark red heat, the torch must be removed. The heat of
the explosion keeps the ball at an almost red heat. The damper
in the air mantle is open, while the starting lamp is burning,
and when closed forms a complete protection. This method of
ignition is extremely simple, reliable and precise.
'^■fSS/ y^ofi U S. Gov£fiM^^fi>r Oct Z OS
A/a
T£ST
LMO
B/tAX£ /^HT/P
MSPOS£A/£ CO//SC/MS£)
\^AT£P
Cof^suMeo
/^e/i/)P/<3
TOTAi.
P£A/y.p.
Paf< /V«/>?
HP/7.
HP.
LBS.
PTS.
LB3.
PTS.
r^J
Les.
/
230
336.7
7S
S8.33
7/.20
.7JS
.SO/
ensiN£ STA/rrws
Mu/vw/vs IOi.e
Z
230
3363
7S
58.0/8
70.74
7343
.895
IC^.
208.6
3
Z30
331.6
79
58.76
7/. 66
7^S
.884-
Am COMP/t£SSOA
WOPHIAJ6 A 6A/MS T
/SO Les Pa£SSUM
IVa,K CCALCuiATia
TOT/iL
8/
NOTES
Load aboot SMP.j Sfieeo 34//1.fiA^ £/v6we /^'^n rfiOM 3SO/l.n to S:SOfin
LOAO 7S H. /'.J 3/=££^o J3S /9 fi M Total. 8 HoufiS^ C/i»f<yify<s /-) Loao of
Lass 'N Spe/d Less Tha/^ /% 7S H P ^r^o no/^£ C^ee Test^SJ.
AiP Oonp/i£3SOfi ■^£''2^' /'u/^PSD UP
T/tMM COA^-TA/^mS ASOUT /O CU rT /f^ ABOUT Tn£ £f</6//^s We/t/<sj> £AJT//>ei.Y
Ze MmUTES /^f>Of^ O TO /so LBS. P^ESSUte. •SATISf^CTOPy Thpoubmout £fj-nn£ fiuN.
Test of 75 H. P. Oil Eugine.
The Method of Operation: The air is drawn into the closed
crank chamber from the interior of the base through a port in
the lower part of the cyHnder. On the forward stroke of the
piston (toward the crank chamber, first stroke) this air is com-
pressed, and a port, opened by the piston, allows it to pass to-
gether with the steam generated in the water jacket to the com-
bustion space of the cylinder. At the same time, the exhaust
port, being overrun and opened by the piston, discharges the
products of combustion. The fuel is injected into the cylinder
by a small pump and there mixed with air and steam so that
on completion of the compression stroke (second stroke) the
mixture of air, oil vapor and steam is automatically fired, the
May, 1912]
MAHER: M. & W. OIL ENGINE
257
258 THE ARMOUR ENGINEER [Vol. 4, No. 2
expansion driving the piston forward and by its connecting
rod, delivers power to the crank shaft.
The cut showing sections of the Mietz & Weiss horizontal
oil engines makes clear its general method of operation. It is
fair to say. that the Mietz & Weiss is the only two-cylinder oil
engine which does not include any experimental ideas. The
difficulty experienced by other manufacturers in having carbon
forming in the cylinders and at the exhaust ports, is overcome
by the use of steam in Mietz & Weiss engines. The following
describes the general method followed :
Cylinders are water jacketed and steam generated passes
through the dome to the air port and, together with the air in
the crank case, to the combustion space of the cylinder, where
it is mixed with oil vapor and exploded at the dead center of
compression. The advantages of this system are: Constant
temperature of the cylinder at varying loads making engine more
rehable ; better lubrication of cylinder ; higher mean pressure
of expansion ; lower mean pressure of compression ; lower oil
consumption ; good fit of piston in cylinder on account of the
even expansion; no water run to waste; reduction of water con-
sumption to a minimum — about two pints per horsepower hour.
The float box, by which amount of water is regulated, it
being impossible for water to run to waste when engine is not
operated, is a simple and standard piece of apparatus. The wa-
ter used in Mietz & Weiss engines does more than scour the
cylinder as it, by partial disassociation, furnishes oxygen for
combustion, making combustion more complete and engine effi-
ciency higher.
The governor is of the fly-wheel centrifugal type, similar
to the high speed steam engine governor. An eccentric on the
main shaft and attached to the governor weight operates by a.
link and rocker arm the oil injection pump. The stroke of this
eccentric becomes less as the governor weight flies out by cen-
trifugal force, which reduces the stroke of the oil pump and
thereby injects less oil into the cylinder. The tension of the gov-
ernor weight spring can be increased or decreased by an adjust-
able screw, to increase or decrease the speed of the engine within
small limits. This spring is adjusted so that at normal speed the
governor weight takes a position between its inner stop at the
fly-wheel hub and its outer stop at the fly-wheel spokes, with-
out, however, striking either. Before this governor weight
touches the outer stop, the stroke of the governor eccentric is
reduced sufficiently to prevent the oil injection entirely because
then the roller at the rocker arm cannot strike the pump plunger
rod. Undue friction of the governor is detrimental to close
May, 1912]
MAKER: M. & W. OIL ENGINE
259
regulation, therefore proper alignment of the governor fly-wheel
with the rocker arm as well as sufficient lubrication of the gov-
ernor weight stud, eccentric, eccentric strap, link, stud and rocker
arm is very essential, because it reduces friction.
The Mietz & Weiss engines of twelve horsepower and above
are equipped with a force feed oiler, very similar to those used
Vertieal Oil E^ngine Air Compressor.
on the better grade of automatic steam engines. The wrist
pin receives oil from the hole in the top of the piston. As the
piston runs back and forward into the cylinder, it passes a hole
in the cylinder wall to which oil is fed from the lubricator by
a sight feed screwed to the pump plate.
The regulation of the Mietz & Weiss engine is very good.
260
THE ARMOUR ENGINEER [Vol. 4, No. 2
/V -<
V f
9 -.
wt?/.
e/A.
nir.
/Ol?
ly^.
£*'/♦
■7be.<?A
\
t
<i
?
\
^
\
!i
\
\
\
s,
\
\
"J
\
V
h)
>• —
J-
\
V
\
!i
\
\
\
\
V
\
\
\
I
\
^P
fj
If}
\
\
.
II
X 1.
'"ife
^
\
q
\
\
\
i
%
I
%
\
^ s
\
^^
\
HJ
"i
1
'
\1
\
'
s>
\
\
\
\
1
\
\
/
1
\
\
\
\
/
\
i
\
\
/
\\
i
1
/
\
\
\
■■.»s
"V-
:..
•O/J.
c/A/
IS'A.
?:?
7/0
7b<i
7a
/
iK
A
§
0
^
J! ^
ri '
\ c
V(
\
!} '
5
5
s
^
5! X
'O.
0.
^// -''*' i/// ^e-iff -'W iy7-/VO/-^c»'/v!^iW<75 .7/i>
May, 1912] MAKER: M. & W. OIL ENGINE 261
Under ordinary load conditions, multi-cylinder engines will keep
within 2 per cent of their rated full-load speed. A large number
of oil engines are direct connected to direct current generators
and without exception have given satisfactory service.
The Mietz & Weiss engines operate on kerosene, distillate
or crude oil. The following is specification of an oil which is
satisfactory for use with all Mietz & Weiss engines:
Fuel oil of specific gravity not exceeding
0.97 with a flash point higher than 600 degrees
Centigrade, with a heat value of not less than
18,000 B. T. U.'s per pound, or any clean
crude oil with a paraffine base, or any kerosene.
The fuel and water consumption curves show in a general
way the efficiency of the Mietz & Weiss engine, and the follow-
ing data shows the guaranteed fuel consumption of a twenty-
five horsepower engine for ten hours at full load, as compared
with a gasoline engine of the same size :
Gasoline
M. & W. Oil Engine. Engine.
Kerosene. Fuel Oil. Crude Oil. Gasoline.
Cents per gallon $ .07 $ .03 $ .02 $ .12
Per day of 10 hours. 1.75 .75 .50 3.00
Per year of 300 days. 525.00 225.00 150.00 900.00
The following shows comparison between eighty horsepower
automatic engine and eighty horsepower Mietz & Weiss oil
engine :
Ten Hours Full Load.
Water at $1.00 per 1,000 cu. ft.
Coal at $3.50 per ton. Oil at 3c per gallon.
Eighty Horsepower Eighty Horsepower
Steam Engine. M. &. W. Oil Engine.
Coal $8.00 Oil $2.50
Water 35 Water 03
Oil and waste 60 Oil and waste 75
Attendant ' 3.00 Attendant 1.00
Total $11.95 Total $4.28
In conclusion, it may be stated that a great deal of interest
is now being taken in oil engines by the best posted consulting
engineers, and that the future of the oil engine business is
doubtless bright, especially for those who have maintained a
record of successful installations.
EXTERNAL FORCES ACTING ON AN AEROPLANE
WHEN IN MOTION.
BY WALTER S. OEHNE.*
An aeroplane while in flight or while in the course of
landing has certain external forces acting upon it. These forces
must be determined so as to be able to figure the stresses in
the different parts of a machine. The methods used in the
succeeding pages are the methods used by students in the course
of Aerodynamics at the Armour Institute of Technology. The
weights and position of the members used in this article are
estimated as closely as possible for a general type of machine.
The type of truss and bracing used, while not referring to any
specific machine, is the general type used in biplanes of the
present design.
Table No. 1 gives the weight of the component parts of a
biplane when ready for a flight. The last column of the table
gives the panel in which the weight is concentrated, the panels
being numbered from left to right as shown in Fig. I.
The propellers and bearings are considered as being sup-
ported on the posts between panels.
Taking the sum of the weights of all the items which are
considered as uniform over the planes, we get 184 pounds.
As there are four cords, 184-^=46 pounds per cord.
As each cord is 39 feet long, 46-^-39=1.18 pounds per foot
of wood.
We want to concentrate this uniform load at the panel
points.
The load from 3 feet of cord is considered as acting at A.
Load at A ■■
= 3
X 1.18 = 3.54 lbs.
Load at a ■■
= 3
X 1.18 = 3.54 lbs.
Load at B ■■
= 6
X 1.18 = 7.08 lbs.
Load at b ■-
= 6
X 1.18= 7.08 lbs.
Load at C
= 6
X 1.18 = 7.08 lbs.
Load at c ■■
= 6
X 1.18 = 7.08 lbs.
Load at Z) = 4.5 X 1.18 = 5.31 lbs.
Load at d =4.SX 1.18 = 5.31 lbs.
To each one of the above loads, one-half the weight of a
post, or one pound, must be added.
The sum of all the weights which are concentrated in panel
4 equals 784 pounds ; one- fourth of this goes to the point D and
♦Class of 1912. Civil Engineering, Armour Institute of Technolog-y.
May, 1912]
OEHNE: AEROPLANE FORCES
263
264
THE ARMOUR ENGINEER
[Vol. 4, No. 2
one-fourth to the point E, the other one-half going to the front
truss, which we are not considering.
The tail weighs 76 pounds, one-fourth of which goes to
each of the panel points d, D, e, E.
The loads which are acting between panels 1 and 2 are con-
Article.
Cloth ...
Sockets .
Clips . . . .
Wire ....
Paint . . .
Seats . . .
Engine
Magneto 12
Oil
Piping
Radiator
Water
Gasoline tank
Gasoline
Two men
TABLE NO. I.
Weight. Panel.
35 Uniform over planes.
10 Uniform over planes.
5 Uniform over planes.
10 Uniform over planes.
5 Uniform over planes.
10 4
215 4
4
4
4
4
4
4
4
4
12
5
35
25
15
75
320
Shafts and brackets and chains.
Propellers (two)
Bearings
Wheels and axles
Rubber bumpers
Tail
Controls and engine frame. ,
Wood used in running gear.
Ribs spaced 1 ft. c. to c. . . .
Posts (a} 2 lbs. each
Spars @ 1.5 lbs. per 6 ft.. ,
80
22
10
30
6
76
20
76
80
32
39
Total weight 1,260
^ in 4
i between 1 and 2
■2 between 1 and 2
-2 between 1 and 2
i between 2 and 3
/2 between 2 and 3
j4 at each panel
point d, D, e, E.
4
^ between 2 and 3
Uniform over planes.
1 lb. to each panel pt.
Uniform over planes.
sidered as concentrated at the lower panel point, or panel point
B, and equal 36 pounds.
The sum of the weight of the wheels, axles, bumpers and
wood which make up the weight of the running gear equals 111
pounds, one- fourth of which goes to the panel point C and one-
May, 1912] OEHNE: AEROPLANE FORCES 265
fourth to panel point F, the other one-half going to the front
truss. Thus the total load concentrated at each panel point is as
follows :
At ^—3.54 + 1 = 4.54 lbs.
At a— 3.54+1 = 4.54 lbs.
At B— 7.08 +1 + 36 = 44.08 lbs.
At h —7.08 + 1 = 8.08 lbs.
At C— 7.08 + 1 + 27.75 = 35.83 lbs.
At c —7.08 + 1 = 8.08 lbs.
At D— 5.31 + 1 + 19 + 196 = 221.31 lbs.
At rf— 5.31 + 1 + 19 = 25.31 lbs.
Landing Stresses.
An aeroplane is stressed when it makes a horizontal turn.
Suppose an aeroplane is going around a circle of radius (r) at
the rate of 45 miles per hour and makes one complete turn in
five seconds. The centrifugal force of a body is given by
il/X4X (3.14)^ r
Where M == mass of the body,
r = radius (in ft.) of the circle about which it
moves.
T = time in seconds which it requires to make one
complete turn.
Mass -- ^v~^g, where zv = weight in pounds.
Substituting in ( 1 ) we get :
_ tc'X4X (3.14)^ r
r = Circumference -f- (2 X 3.14)
45 mi/hr = 66 ft/sec
Therefore,
66 X 5
= 52ft.
2X3.14
1260 X4X (3.14)^X52
32.2 X 5 X 5
2900 lbs.
This force acts horizontally while the weight acts vertically.
Therefore, resultant is equal to the square root of [(2900)^ +
(1260)2] ^hich equals 3160 pounds.
Each pound of weight of the machine exerts a force of
3160-^-1260 = 2.5 pounds, parallel to the resultant.
266 THE ARMOUR ENGINEER [Vol. 4, No. 2
Multiplying each of the loads at the panel points by the con-
stant 2.5 it will give the force at the respective panel points due
to the loads concentrated there.
In order to have these forces act vertically the machine is
considered as being in a horizontal position.
This gives the following forces :
A— 4.54X2.5= il.35 lbs.
a^ 4.54X2.5= 11.35 lbs.
5_ 44.08X2.5 =111.2 lbs.
b— 8.08X2.5= 20.2 lbs.
C— 35.83 X 2.5 = 89.61 lbs.
c— 8.08X2.5= 20.2 lbs.
Z)— 221.31 X 2.5 -= 553.27 lbs.
d— 25.31 X 2.5 = 63.27 lbs.
The sum of these multiplied by 2 gives the total downward
force on the rear truss or 2 X 880.45 = 1760.9 pounds, and this
must be balanced by the pressure of the air acting on the planes
in the opposite direction. Or,
39X45X>^X2
'■ — = 10 lbs. per sq. ft. of planes.
1760.9
Multiplying the area of planes which is considered as car-
rying any pressure on it to its respective panel point by the con-
stant 10, we get the total upward pressure which is taken as
acting at the panel point, or at
A—{Z X 4.5) X >^ X 10 = 67.5 lbs.
a— 67.5 lbs.
5— (6 X 4.5) X >4 X 10= 135 lbs.
&— 135 lbs.
C— 135 lbs.
c— 135 lbs.
D— 4.5 X 4.5 X >^ X 10= 101.25 lbs.
d— 101.25 lbs.
Calling the downward forces we obtained by multiplying
the loads by the constant 2.5 negative, and the upward forces
just obtained positive, we get a resultant force at each panel
point of
^_ 67.5 — 11.35 = + 56.15 lbs.
a_ 67.5 — 11.35 = -}- 56.15 lbs.
5—135 —111.2 =+ 23.8 lbs.
&— 135 — 20.2 = + 114.8 lbs.
C— 135 — 89.6 =+ 45.4 lbs.
c— 135 — 20.2 =+114.8 lbs.
Z)— 101.25 — 553.27 = — 452.02 lbs.
rf— 101.25— 63.27 = + 37.98 lbs.
May, 1912] OEHNE: AEROPLANE FORCES " 267
The algebraic sum of the above should equal zero. Any
slight error is due to dropping of the last decimal place through-
out parts of the work. This error can be distributed so that the
algebraic sum will equal zero.
It is the above loads just obtained which are used to fig-
ure the stresses in the members. The stresses are figured the
same as in any ordinary truss of the type shown in Fig. 1.
The only thing to be remembered is that the diagonals in aero-
plane trusses are made of wire and therefore taken tension only.
Horizontal Turns.
A favorite pastime of aviators while flying is to glide down-
ward, either with their engine running or shut ofif, and, then
to suddenly turn the machine upward. This is a vertical turn
and it puts a considerable stress on the members of the machine.
In a typical example an aeroplane approaches the ground
at the rate of 90 miles per hour and then makes a turn up-
wards, this turn having a radius of 100 feet.
Using the same symbols as before, the centrifugal force
is given by
mir Wv-
r gr
where v is the velocity in feet per second.
Substituting the values :
1260X132^
f = = 6818 lbs.
^ 32.2 X 100
Reasoning the same as in the case of horizontal turns,
6818^1260=5.41 pounds downward force exerted by each
pound of weight.
Downward forces at each panel point are:
A— 4.54 X 5.41 = 24.56 lbs.
Or- 4.54 X 5.41 = 24.56 lbs.
B— 44.08 X 5.41 = 238.47 lbs.
b— 8.08 X 5.41 = 43.71 lbs.
C— 35.61 X 5.41 = 192.65 lbs.
c— 8.08 X 5.41 = 43.71 lbs.
Z}— 221.21 X 5.41 = 1197.39 lbs.
d— 25.31 X 5.41 = 136.93 lbs.
Or the total downward pressure on the truss equals twice
the sum of the above, or 2 X 1901.95 = 3803.9 pounds, which
268 THE ARMOUR ENGINEER [Vol. 4, No. 2
must be balanced by the air pressure in the opposite direction.
Therefore
39X4.5X>4X2 ^^ ^ ,^
^21.7 lbs. per sq. ft.
3803.9
The force at each panel point due to the air pressure on the
planes is considered as acting upward and is equal to :
At A— 3 X 4.5 X >4 X 21.7 = 146.47 lbs.
At a— 146.47 lbs.
At B—6 X 4.5 X >< X 21.7 = 292.95 lbs.
At &— 292.95 lbs.
At C— 292.95 lbs.
At c— 292.95 lbs.
At D-A.5 X 4.5 X K X 21.7 = 219.71
At rf— 219.71 lbs.
Calling the downward forces negative and the upward
positive as before we get a resultant force at each panel point of
^ = 146.47 —
24.56 =
= 4-121.91 lbs.
0=146.47 —
24.56 =
= + 121.91 lbs.
B = 292.95 —
238.47 =
= + 54.48 lbs.
b = 292.95 —
43.71 =
= + 249.24 lbs.
C = 292.95 —
192.65 =
= + 100.3 lbs.
c = 292.95 —
43.71 =
= + 249.24 lbs.
D = 219.71 —
1197.39 =
= —977.68 lbs.
rf = 219.71 —
- 136.93 =
+ 82.78 lbs.
And as before the algebraic sum should equal zero. These
are the loads used to figure the stresses.
Gilding Forces
If a machine is just gliding through the air and not making
any turns such as a horizontal or vertical turn it is acted on
by certain external forces. If we divide the final loads at the
panel points which we got for a machine while making a hori-
zontal turn by the constant 2.5, which is the constant we multi-
plied each load by in that analysis, we get the load at each panel
point which is used to figure the stresses for gliding only. Or
if we divide the the loads gotten by the analysis the vertical
turn by the constant used in that case (5.41) we would also
get the loads at the panel points from which the stresses are
figured for gliding only.
Vertical Turns.
If we have a machine approaching the ground at the rate
of 40 miles per hour making a slope of one to seven ; that is,
one vertical to seven horizontal, the instant it hits the ground
May, 1912] OEHXE: AEROPLANE FORCES 269
it will be acted upon by certain forces. On the machine used
in this work we will assume that the springs which are on the
wheels or running gear and which come into play when the ma-
chine touches the ground are compressed six inches ; that is, an
instant before the machine touches the ground the springs are
in their normal position, while an instant after the machine
touches the ground they are compressed six inches.
K.E. =
2g
where v is the vertical component of the velocity.
Average pressure of the springs :
2(j 2gs
where s is the distance through which the springs act in feet, and
F is the average pressure. The maximum pressure on the
springs when compressed or extended will be equal to 2F, be-
cause the pressure varies from zero to a maximum.
Therefore,
2F =
9^
The vertical component of 40 miles per hour equals 8.3 feet per
second.
1260 X 8.3-
2F = = 2696 lbs.
32.2 X -5
2696
1260
14
Multiplying each of the loads at the panel points by the constant
2.14, we get the external forces acting on the machine at the re-
spective panel points as follows :
A —
4.54 X 2.14 =
9.72 lbs.
a —
4.54 X 2.14 =
9.72 lbs.
B —
44.08 X2.14 =
94.33 lbs.
b —
8.08 X 2.14 =
17.29 lbs.
C —
35.61 X 2.14 =
76.21 lbs.
c —
8.08 X 2.14 =
17.29 lbs.
D —
221.31 X 2.14 =
473.60 lbs.
d —
25.31 X2.14 =
54.16 lbs.
The upward pressure of the air on the planes is not con-
sidered in the above analysis, because a machine on rough ground
270 THE ARMOUR ENGINEER [Vol. 4, No. 2
is apt to fully extend its springs a number of times before com-
ing to rest, and this is liable to happen at such low velocities
that the pressure of the air on the planes is negligible.
Running Along the Ground
There are certain forces which act on a machine when it
runs along the ground, either before it starts on a flight or when
it comes to rest, after a flight. Using the same data as before,
that is, a machine moving toward the ground at the rate of
40 miles an hour and at a slope of one to seven, we will assume
that it takes 150 feet for the machine to stop after it strikes the
ground. Horizontal component of 40 miles an hour equals 58
feet per second.
KE = y2 mz'- = y2 zm'- -^ g ^ Fs.
Wv-
Fs =
Therefore,
438.8 lbs.
F =
2gs
1260 X 58^
F =
2 X 32.2 X 150
438.8
= .349,
1260
which is the constant by which the load at each panel point is
multiplied so as to give the horizontal forces at the panel points
which should be used to figure either the bending stresses in the
chords or the stresses in the horizontal bracing if there are any.
At A —
4.54 X
.349 =
1.58 lbs.
At a —
4.54 X
.349 =
1.58 lbs.
At B —
44.08 X
.349 =
15.38 lbs.
Atb —
8.08 X
.349 =
2.82 lbs.
At C —
35.61 X
.349 =
12.43 lbs.
At c —
8.08 X
.349 =
2.82 lbs.
At D —
221.31 X
.349 =
77.24 lbs.
At d —
25.31 X
.349 =
9.23 lbs.
There is one set of stresses which should yet be determined
and those are the stresses which are due only to the weight of the
machine. These can be obtained by dividing the stresses gotten
in the case of "landing" by the constant 2.14.
A TALK ON THE PHILIPPINE ISLANDS.f
BY E. L. LUNDGREN, C. E.*
The Philippines are a group of over three thousand islands
southeast of the Continent of Asia. They lie (between 116° 40'
and 126° 34' east of longitude and between 4° 40' and 21° 10'
north latitude) wholly within the tropics and almost half way
around the world from Chicago.
The commercial route to the Philippines is by the way of
Japan and China and is only fifteen hundred miles longer than
the most direct route. On account of loading and unloading
freight, stopovers are made in Japan of five to ten days and in
Shanghai and Hong Kong of one to three days, the duration of
the journey from Chicago being about thirty-five days. The
business men of the Philippines are agitating for a hne of steam-
ers to cross directly from San Francisco to Manila. If this could
be secured the trip could then be made from Chicago in less than
three weeks.
The total area of the Philippines is about 115,000 square
miles. This is approximate only, as no accurate map is in exist-
ence. The American and Philippine governments are jointly
engaged in coast and geodetic work and have charted slightly
over half of the islands to date. Strange to say, the coast line
of the Philippines is about double that of the mainland of the
United States.
The population of the Philippines is over eight million
people, of whom seven and one-quarter million are classed as
Christian and the balance as Mohammedans and pagans. The
population in the year 1570 was only half a million and in the
year 1800 was a million and a half. Out of the eight million
people only thirteen thousand are white, many of whom are Eng-
lish, Spanish and German. The United States army is excluded
from this number.
The area and population of the Philippines are about equal
to the area and population of Illinois and Wisconsin combined.
The difference in economic conditions, however, is very striking.
The Filipinos are classed as an agricultural people, but they cul-
tivate only five per cent of their area, while we in IlUnois and
Wisconsin with a large percentage of our people engaged in
manufacturing cultivate sixty per cent of our area. The rapid
♦Class of 1904. Project Engineer, Bureau of Public Works, United States
Government, Manila, P. I.
tDelivered before the M^estern Society of Ensrineers, May 6, 1912.
272
THE ARMOUR ENGINEER [Vol. 4, No. 2
growth of vegetation, the antique methods of agriculture and the
legarthic condition of the people, due to living in the tropics, all
tend towards an extremely small farm unit. A family of Fili-
pinos will cultivate from two to five acres, while here a family
will cultivate eighty to one hundred and sixty acres. The Philip-
pine government is making every effort to introduce modern
methods of agriculture. In addition to experimental farms and
demonstrations of American machinery, the schools devote a
large portion of their curriculum to agricultural and industrial
work.
The people very seldom live on their farms, but congregate
in villages for the sake of companionship and mutual protection.
Te.st Boring Rig on Propo-sert Tunnel. Vgno River Project, Province of
Pangasinan.
During the rice harvest men, women and children all turn out
to work. According to Filipino custom each head of rice is cut
separately by hand and tied in bundles, the pickers receiving one
bundle out of every four for their labor.
The political organization of the Philippines is similar to
that of our government. The executive is called the governor
general, the upper house the commission and the lower house
the assembly. The lower house is composed entirely of Filipinos
elected to office by Filipino voters. The upper house is composed
of the governor general, who is the chairman, and eight com-
missioners, four of whom are Americans and four are Filipinos.
All of the members of the commission are appointed to office by
the President of the United States. Three of the American and
May, 1912] LUNDGREN: THE PHILIPPINES 273
one of the Filipino commissioners also act as secretaries of de-
partments and receive extra compensation for this work. The
entire legislative body of the Philippines is composed of natives
with the exception of the governor general and four commis-
sioners.
In a recent speech before the Chicago Commercial Club Mr.
Quezon, Philippine delegate to Congress, in his plea for imme-
diate independence, speaks on the injustice of the large salaries
paid to the commissioners in the Philippines. He neglects to
state that half of them are Filipinos, who evidently feel that
they earn the money, as I have not heard of any of them return-
ing a portion of it to the government. He also fails to comment
on the fact that the Philippine assembly when elected to office
immediately increased their salary from ten to fifteen dollars
a day.
President McKinley announced that his Philippine policy
was to have the Americans govern the islands until the time
came when the Filipinos were able to govern themselves. This
policy is very humane and logical, but has stirred up a storm of
protest from the large majority of the Filipinos as well as a
large minority of the Americans. The Filipino says that he is
now ready to govern himself — that he is afraid that our announced
policy is a subterfuge and that we intend, eventually, to annex
the islands to the United States. He feels that he is entitled to
the offices that the Americans now enjoy and that he could
change conditions to the better advantage of the Filipinos. The
American says that we not only conquered the islands from the
Spaniards, but also from the Filipinos, and that in addition have
purchased them from Spain for twenty million dollars and have
sunk hundreds of millions of dollars into the country since the
purchase; therefore it is only logical that the Americans be
allowed to exploit the country for their own benefit.
The American administration in spite of these attacks had
adhered strictly to their instructions as formulated by our late
President McKinley. The elective franchise has been rapidly
extended to the Filipino voter. First the municipal governments
have been turned over to their charge, then the provincial gov-
ernments and next the lower house of the legislature. The only
steps that remain before entire self-government is the power to
elect their upper house and their executive.
When this will be done is entirely a matter of conjecture.
President McKinley promised it to the Filipinos when they could
govern themselves. In the civil service report of the Philippine
Islands for 1911 I find there were 3,307 Americans and 4,023
Filipinois employed in 1905, and in 1911 there were 2,633 Ameri-
cans and 4,981 Filipinos in the service. This shows a decrease of
274
THE ARMOUR ENGINEER
[Vol. 4, No. 2
664 Americans and an increase of 958 Filipinos. At this rate the
entire service should be Filipinized in twenty-five years. This
agrees w^ith a statement made by President Taft as to the time
that should elapse before the Filipinos should secure their inde-
pendence. Three solutions are possible to the Philippine ques-
tion, namely, permanent annexation to the United States, imme-
diate independence and deferred independence.
Annexation is favored by those Americans who think that
we, like the larger European nations, should have a trade center
near Asia to aid the expansion of our commerce, and in addition,
Scene on the Beunuet Kt>a<i \ii<-i '*<orin.
During this typhoon the world's record for rainfall was
broken; 34.5 inches in one day, and 88 inches in four days, at
Baguio, Province of Benguet, Philippines.
that the islands could be exploited in a manner similar to the
Hawaiian Islands, so that the United States would be compen-
sated for the enormous expenditures we have made in the Phil-
ippines.
Immediate independence is favored by a vast majority of
the Filipinos as a national instinct. No race cares to be governed
by another race, no matter how paternal the government. It is
also favored by a large percentage of Americans, who point out
that the cost of the Philippines to the American taxpayer is
between ten and fifteen million dollars per year in addition to
what it would cost if all of our army was kept in America. They
claim that our manufacturers cannot supply the markets on our
May, 1912] LUNDGREN: THE PHILIPPINES 27:
own continent for many decades and that the Philippines would
be a source of weakness in time of war.
Deferred independence is the present policy of our govern-
ment and will continue to be unless the American voter deter-
mines otherwise. The governor general and the commissioners
in view of the opposition of those who do not agree with their
instructions deserve great credit for the tact and skill with which
they are doing their work.
In consequence of this policy, the Philippines have a com-
plete government organization which is distinct from that of the
Method of Throwing- Up E^arth Embankments in Swampy Ground.
United States. Their laws, postage stamps and currency are
entirely different from ours.
The Philippine civil service employs a large number of
Americans in various capacities, but they are being replaced as
rapidly as Filipinos can be educated for the offices. Ultimately
all of the Americans will be replaced by Filipinos, so when they
receive their independence there will be no confusion in the
government due to Americans leaving the country.
The government shortly after it was reorganized by the
Americans placed the currency of the country on a gold basis,
arbitrarily making the peso equal to half of an American dollar.
The coins issued are similar to ours, but are worth only half as
much. They are the centavo (a copper coin about twice the size
of our cent), the five centavo (a nickel coin the exact size of our
five cent piece), the major peseta (a silver coin similar to our
276
THE ARMOUR ENGINEER
[Vol. 4, No. 2
dime), the peseta or twenty centavo piece (this fractional silver
coin is more convenient in making change than our twenty-five
cent piece), the major pesos or fifty centavo piece and the peso,
or one hundred centavo piece (silver coins similar to our half
dollar and dollar). Paper currency only is issued for money
above a peso. These bills are much smaller than our American
notes, being only two-thirds as long and two-thirds as wide.
This smaller size is found more convenient to handle than the
larger American bills.
The only postage stamps that can be used in the Philippines
Typical Cart in Present Use in the Philippines.
are those issued by the Philippine government, which fact does
not seem to be understood in the United States. Practically all
who have dealings in the Philippines send their addressed return
envelopes and cards with American stamps on them, losing thou-
sands of dollars annually from this item alone. Much better
results would be obtained if Filipino stamps were placed on the
return cards or else a statement of the denomination of the
Filipino stamp necessary to mail the return card.
The Philippine government collects the customs and internal
revenue in the islands and the proceeds go into the Philippine
treasury. The United States receives no revenue from the
islands.
May, 1912] LUNDGREN: THE PHILIPPINES 277
The engineering work in the Philippines is mainly under
the supervision of the Bureau of Public Works. This bureau
has charge of the construction of roads, bridges, public buildings,
artesian wells, irrigation systems and river control works except
in the city of Manila, which has its own engineering department.
The bureau of navigation has control of the construction of port
works and lighthouses.
During the last four years the personnel of engineers of the
Bureau of Public Works has increased from fifty to one hundred
and ten and its expenditures from one million to five millions
dollars per year. The average length of service of the engineers
is about three years and the average annual salary is about two
thousand dollars. On June 30, 1911, the total number of em-
ployees was 1,489, an increase of twenty-seven per cent over the
preceding year. Of this number 564 were Americans and 925
Filipinos. These figures do not include laborers and native
foremen hired on force account.
Over three million dollars is now spent annually on public
roads. On account of climatic and traffic conditions the de-
terioration of a high grade macadam road in the Philippines is
exceedingly rapid, the rate of deterioration being about twenty-
five per cent per year. In other words, four years after a road
has been constructed it is practically impossible for wheeled
traffic. The Spaniards were good road builders and the remains
of their work can be seen all over the islands. The United States
army, while it administered the country and later the Philippine
government as organized by the Americans, spent large sums in
road construction, but as fast as new roads were built the old
ones went to pieces. Sledges were in common use and on most
of the carts the wheels were narrow rimmed and were fixed
solidly on the axle.
(The accompanying illustration shows a typical cart as used
in the Philippines today. In Spanish times the wheels were a
solid section cut out of wood and set solid to the axle, with
very narrow rims; the Americans have taught the natives how
to manufacture cart-wheels with broad rims, such as are used
in the United States. Chinese coolies are shown unloading
sugar molds. The carabas or water bufifalo is harnessed to the
cart by a yoke and is guided by a rope tied through the nostrils ;
a long, steady pull on the rope turns him to the right and a
series of jerks turns him to the left.)
In 1908 a new road policy was inaugurated. Sledges and
narrow rimmed vehicles were not allowed on improved roads.
The two hundred and thirty-seven miles of good road that then
existed were put under maintenance. One laborer working con-
stantly on each half mile was found to be necessary to keep the
278 THE ARMOUR ENGINEER [Vol. 4, No. 2
road in good condition. This work consists of filling up ruts
with surfacing material, cutting down tropical growth, cleaning
side ditches, etc. In addition, it is necessary to completely resur-
face the road every five years. The expense of this maintenance
has been found to be about five hundred dollars per mile per year,
and as the cost of construction varies from five to ten thousand
dollars per mile, the deterioration of improved roads under the
present system, as measured in cost of upkeep, is only from five
to ten per cent as compared with twenty-five per cent under for-
mer conditions. Since this system has gone into efifect the mile-
age of first class roads has increased from two hundred and
thirty-seven miles in 1908 to nine hundred and fifty miles in 1911.
The Bureau of Public Works has completed plans to con-
struct a road from one end of Luzon to the other and hopes to
accomplish this important work in the next two years. A very
thorough study has been made of the roads needed for the eco-
nomic development of the Philippines, From this study pro-
posed road construction has been laid out so as to build roads
that will benefit the greatest number of people. By this I do not
mean that roads are to be built only in congested areas, to the
neglect of the more sparsely populated districts, but it is the
intention to spend the money in a direct ratio with the density
of the population throughout the Philippines. This has led to
the construction of a large number of disconnected stretches of
road, generally out of each provincial capital, which is usually
the center of commerce for the province. All work is being done
with the complete road system in view, and no deviation from
this plan is permitted.
Wood, on account of its rapid decay in the tropics, has been
abandoned for bridge construction and reinforced concrete is
now used almost exclusively.
A district engineer is stationed in each province and has
charge of the construction of the public buildings and roads in
that province. He may have one or more assistant engineers,
depending on the quantity of work to be done. Most of the
culverts and roadways are built by force account, while the
bridges and public buildings are mostly constructed by contract
under the supervision of the district engineer.
The various provinces are combined into five divisions, each
under a division engineer, who inspects the work of the district
engineer and sees that construction is kept up to the standard
set by the chief engineer of the bureau.
Irrigation and river control work is executed by another
division of the Bureau of Public Works. The work is done under
the immediate supervision of a project engineer, who reports to
the chief irrigation engineer of the bureau. The funds for irri-
May, 1912]
LUNDGREN: THE PHILIPPINES
279
gation work are provided by special act and the plan is to have
the benefited parties reimburse the government for the work
done.
The irrigation division was organized in 1908. The engi-
neers found that nothing was known as to hydrographic condi-
tions and that no contour maps were in existence. First various
projects were roug-hly blocked out, gaging stations established
and reports with approximate estimates were made. From these
reports it was decided whether the project should be abandoned
or whether detailed surveys should be made. A few small
Ga^in^ Station on the Agno River; Span
projects were found to be so favorably situated that construction
was begun immediately after the surveys were completed, but on
the larger projects the collection of hydrographic data for at
least five years was deemed essential.
(The accompanying illustration shows a gaging station on
the Agno River. The span is 680 feet ; the main cable is three-
quarters of an inch in diameter, while the tag and stay lines are
three-eighths inch. The more vertical leg of the A-frame is the
trunk of a tree that was growing on the spot. Gagings were
taken by an American hydrographer who visited the station
weekly, while a Filipino gage-keeper lived in the vicinity and
read the river and rain gages twice a day.)
A vast quantity of information has been collected to date.
280
THE ARMOUR ENGINEER
[Vol. 4, No, 2
More than two hundred projects have been considered, of which
over half have been abandoned as not feasible. Final construc-
tion surveys have been made on five projects and construction
is expected to commence on the completion of the hydrographic
studies.
Trail >Ia«le by Anierioau Ku^^iiieerx Over the Mountains.
The rapidity of growth of vegetation may be judged by the
length of the grass in the foreground, which was cut one
month previous.
An artesian water supply was discovered about five years
ago and the drilling of wells for municipal water supply has
become an important branch of the Bureau of Public Work.s. The
benefit of improved drinking water is shown by a rapid decrease
in the death rate wherever an artesian supply is secured.
About one hundred and twenty miles of railway were in
May, 1912]
LUNDGREN: THE PHILIPPINES
281
Operation when the United States conquered the Phihppines.
The Americans have made successful efforts to increase this
mileage; 414 miles were in ooeration in 1908 and 585 miles in
1911, an increase in three years of 171 miles. Six hundred and
ninety-eight miles of proposed construction have been authorized
Bridge on the Benguet Road.
by the Philippine government and ultimately the islands will
have a very complete system of railways.
The friar-lands question has received a great deal of atten-
tion from Congress, and has become a matter of interest. The
friars consist of a number of orders affiliated with, but at the
same time, entirely distinct from the organization of the Catholic
Church. At first these orders gave great assistance to the Span-
ish government and were the prime factor in civilizing the Fill-
282
THE ARMOUR ENGINEER
[Vol. 4, No. 2
pinos. Naturally they gained the respect and good will of both
the natives and the government and they were deferred to by all.
Unfortunately this taste of power was demoralizing and the
orders became greedy for more power and luxury. They became
so powerful that they could unseat the governor general at will.
Conditions became so bad that the Spanish king expelled the
Jesuits and only allowed their return many years later under
very restricting conditions. At the time of the American occu-
pation the four orders of friars who were allowed to possess
property were the Augustinian, having 346 friars in the Philip-
pines ; the Franciscan, having 107 friars ; the Dominican, having
aw
mt^p^m
ICi^^f
m ^S'-m,. iH^*
*^Kmj'
wa^r^mM
5
h-' '
Interior View of Wall Around the Spanisli City of Manila, Showing Gate
and Dun!a;eon.s in Wall.
233 friars, and the Recoletos, having 327 friars. They were able
to show title to 420,000 acres of land.
The friars are said to have acquired title to these enormous
holdings in a questionable manner. This matter was a source
of constant friction between the Spaniards and the Filipinos and
appears to have been one of the main causes of the various
insurrections against the Spaniards. In order to settle the mat-
ter Congress authorized the Philippine government to purchase
the land for the purpose o,f resale to actual residents; 410,000
acres out of the 420,000 acres were purchased for $7,239,000.
Practically all of the cultivated area has now been disposed of.
A large portion of the land, however, was found to be unsettled
May, 1912]
LUNDGREN: THE PHILIPPIXES
^83
and when they were unable to sell it to individuals the govern-
ment began to sell the land to corporations in large tracts. Con-
gress was informed of this and stopped the sales. The Philip-
pine government has this property on hand and is unable to sell
it as Congress desires, so they now advance the argument that
the American government should pay for this land in order that
the bonds may be retired.
The friar lands should not be confused with the lands of the
Catholic Church. The church also has enormous holdings, but
they were not purchased by the government.
Tourists arriving in Manila will find it one of the most
street Scene in Manila, P.
Bureau of Printing in Baek^rouncl.
interesting cities in the world. In modernizing the city the Amer-
icans have made every effort to enhance the beauty of the old
mediaeval city they found instead of destroying the old scenes,
as is usually done.
Mr. D. H. Burnham, of Chicago, was called to the Philip-
pines by the government soon after it was organized and drew
up the plans for future improvement of Manila on the same
principle as the proposed city plan of Chicago. While there
he also visited the site of the proposed summer capital at Baguio
and drew up a tentative plan for construction of that city.
One of the most interesting features was his treatment of
the old walled city. Paris replaced her walls by a beautiful
boulevard, and in one way or another most of the walled cities
284 THE ARMOUR ENGINEER [Vol. 4, No. 2
of the .world have disappeared. Manila, however, has hers intact
and the space formerly occupied by the moat has been filled in
and made into sunken lawns similar in appearance to our Midway
Plaisance. This not only brings out the full effect of the old
fortifications, but creates a large playground, where hundreds of
Americans and Filipinos play baseball, association football and
other games in the afternoon, attracting thousands of spectators.
The harbor problem was solved by building a new break-
water and dredging a new basin, the excavated soil being used to
fill in the old moat and for making new land along the water
front. The new fill is used mainly for warehouse and harbor
facilities.
Another interesting feature of the Spanish occupation is the
Luneta. This is a small park and driveway where the people of
Manila assemble in the evenings and listen to the music of vari-
ous mihtary bands. This glimpse of Manila on dress parade is a
sight that I do not think is ever forgotten by one who has seen
it. To give this feature a better setting and to provide space for
the government center, Mr. Burnham planned to have it moved
to the new fill.
INDUSTRIAL ELECTRIC SERVICE IN THE
SOUTHERN STATES.f
BY A. D. QUACKENBUSH.*
In reaching out after some of the large industrial establish-
ments located in the environs of Mobile, Alabama, the Mobile
Electric Company found it expedient to erect 11, 000- volt lines
from the power house to connect up a number of factories hav-
ing a load of over 1,110 hp. The motor equipment of two of
these establishments is described in detail herewith as well as the
i
eimM
ilii^^B^^I
FlK- 1. Substation, WTilstler, Ala.
construction of the 11,000-volt lines over which the energy is
transmitted.
The 11,000- Volt Line.
There are two distinct 11,000-volt lines leaving the power
house, one going to Whistler and the other to the Dauphin Lum-
ber Company. The former consists of three No. 3 medium hard-
drawn copper wires, which are insulated within the city and
bare outside the city limits. These wires are spaced on thirty-
inch centers, in the form of an inverted triangle, and are mounted
on the same side of the pole on four-inch by five-inch cross-arms.
This construction allows a duplicate line to be built on the other
*Class of 1907. Assistant Superintendent, Mobile Electric Company,
Mobile, Alabama.
tPublished in the Electrical World, March 2, 1912.
286
THE ARMOUR ENGINEER
[Vol. 4, No. 2
side of the pole and leaves the top of the pole vacant for a
ground wire. The pole line runs along the Mobile & Ohio Rail-
road right of way. Much difficulty was encountered in setting
some of the poles owing to the swamps. Many poles had to be
set in barrels and had to have four poles used as braces to keep
them afloat. All poles are juniper; cross-arms are of longleaf
yellow pine; pins within the city are of locust and outside the
city are of oak.
In the early spring many of the oak pins began to break off,
causing the insulators to turn upside down on the cross-arm.
The pins broke off just inside the inner petticoat of the insu-
lators. The wood looked rotten and was in a rather spongy con-
Fi;;. 2. Hlsb-Tentilon Bus, Whistler Substation.
dition, but not charred. The locust pins were in perfect condi-
tion. In order to prevent any more trouble from these pins,
they were sawed off and galvanized iron bracket pins which
clamp around the arms were installed in the same position as the
former pins. As the work had to be done on Sundays, the
expense of installing these pins amounted to twenty-two cents
each.
The electrical storms around Mobile are frequent and the
lightning is severe, but the lightning arresters at each end of the
line have proved very efficient. Many times the resistance rods
of the arresters have broken, and only once has the circuit been
knocked out. One of the poles had the entire top knocked off
by lightning. On a portion of the line about a mile long a No. 4
galvanized steel wire has been strung on the top of the pole for
May, 1912]
QUACKENBUSH: ELECTRIC SERVICE
287
a guard wire. Whether this has any advantage on this particular
Hne is open to discussion.
About three miles from the power house the first branch
line taps the main line through disconnecting switches. This
line is about half a mile long, and has four 40-kw. transformers
FlK-. 3. Llisrhtnlng Arresters in ^Vhistler Substation.
connected to it. Three 40-kw. transformers were installed for
the Mobile abattoir, which has a connected load of 67^ hp.
These transformers are mounted on a wood platform between
poles and are protected by General Electric ll,0(X)-volt fuses.
The single 40-kw. transformer steps the voltage down from
288
THE ARxMOUR ENGINEER
[Vol. 4, No. 2
11,000 to 2,300 and is used to furnish a primary energy for the
Hghting of Pritchard.
The next branch line is about three-fourths of a mile long
and furnishes two power consumers. The first one has three
40-kw. transformers installed on a platform and operates a plan-
ing mill ; the second installation consists of three 10-kw. trans-
formers on a pole. It operates a stave mill.
The third branch taps the line in Whistler and operates a
chair factory about a mile distant. The installation consists of
three 50-kw.' transformers protected by Westinghouse high-
FicT- 4. Temporar>' TranHforiners for Dauphin Lumber Company.
tension fuses and mounted on a platform between two poles.
The total connected load is 145 hp.
The line going to the Dauphin Lumber Company consists of
three No. 6 medium hard-drawn copper wires mounted on porce-
lain insulators, all on the same arm. The one-arm construction
was used because of its cheaper first cost. Should a duplicate
line ever be built, a second arm can be installed. The triangular
spacing of the wires for this voltage and short distance ( six
miles) shows no advantages over placing all the wires on the
same arm.
Power-House Equipment.
When the Whistler line was built three 1 50-kw. General
Electric water-cooled transformers, stepping the voltage from
Alay, 1912] QUACKEXBUSH: ELECTRIC SERVICE
289
2,300 volts to 11,000 volts were installed. Multiplex lightning
arresters in connection with choke coils protected the transform-
ers. The bank is controlled on the 2,300-volt side and is pro-
tected by an overload relay. When the Dauphin Lumber Com-
pany contract was received a duplicate equipment was installed.
In addition a 15,000-volt hand-operated, non-automatic oil switch
w-as installed in each bank. On the line side of these switches
are disconnecting switches, and on the line side of the disconnect-
ing switches is a 15,000-volt bus tie switch protected on each side
by disconnecting switches. By this arrangement a bank can oper-
K'r l.i
0^'
I
mmm
■i
"*;'
Fie. .'t. Sab8tatlon for the Dauphin Lumber Company.
ate its respective line, a bank can operate the two lines, or the
banks can be paralleled on the two lines.
Shops of the Mobile & Ohio Railroad.
The shops of the Mobile & Ohio Railroad are located in
Whistler about six miles north of Mobile. As these shops were
in operation during the civil war, they are very old, and various
machines were added from time to time until the steam equip-
ment became inadequate and in order to carry on operation one
department would have to stop work so as to have sufficient
power to operate the machines in another department. Instead
of installing a new steam plant an electric drive was established
in the spring of 1908. The cost of the change from steam to
290
THE ARMOUR ENGINEER
[Vol. 4, No. 2
electricity was very large, so it was decided to install a group
drive.
There are thirty-eight motors in the shops, all of three-phase,
440-volt Westinghouse induction type, aggregating 588^ hp.
The shops operate nine hours a day, but in a few cases have
operated twenty-four hours a day. Based on a twenty-four hour
day and twenty-six working days a month, the load factor is
sixteen per cent. The average number of kilowatt hours used
a month is 44,705.
The substation is located in the center of the longest of
Fler. e. view o« Traiuimlsslon Line Near Whistler.
several buildings. This particular point used to be the entrance
to a boiler room, and as box cars have to come in this entrance
the floor of the substation was built fourteen feet above the
ground. In the station there are three 100-kw., 10,500-440-volt
water-cooled transformers, each separated by a four-inch rein-
forced concrete barrier. The line enters the substation through
sewer tiles built in the wall, and after passing through choke
coils and series instrument transformers enters a 15, 000- volt
switch located in a concrete compartment. The buses then go
direct to the transformers. The only instruments used are a
May, 1912] QUACKENBUSH: ELECTRIC SERVICE
291
kilowatt-hour meter and a voltmeter. The transformers are pro-
tected by General Electric multiplex Hghtning arresters.
The Dauphin Lumber Company.
The next large and notable installation is that of the Dau-
phin Lumber Company. This company always had its own
A
fU
"^
f
1
1
* i^
^
T^S
u
fm
\
■■'
r'V
'?^
1 ^ .^^
'^BB^^^Si^f'i^^S^
lg
ilMi
FIc. 7. Transformers for Adier May Company.
plant, and each machine was either direct-connected or belted to a
direct-current compound-wound 220-volt motor.
The work done by the Dauphin Lumber Company consists
in dressing rough lumber for export. The company sometimes
enters into a contract to ship a given number of thousands of
feet of lumber within a specified time. On the night of May
10, 1911, the power plant burned, thus placing the mill out of
commission. Something had to be done immediately, as the
292
THE ARMOUR ENGINEER
fVol. 4, No. 2
company was under heavy demurrage. A contract was signed
with the Mobile Electric Company on May 12, 1911, to have
the mill in operation by May 16, 1911. In order to do this all
the direct-current motors had to be removed, and the mill had
to be rewired, as three-phase alternating current was to be used
instead of direct current. Some of the motors were connected
FlR. 8. Trnnsformer Installation for Corinth Chair Factory.
by means of flexible coupling to the machines, while others were
connected by rigid couplings, and in order to change these it
was necessary to bore some of these couplings and bush others
so that they would fit the shafts of the new motors.
The nearest three-phase line was about half a mile distant.
This extension was made and nine 20-kw. transformers were
installed, connected delta, using three transformers in multiple
per phase. As the Dauphin Lumber Company is located nearly
May, 1912] QUACKENBUSH: ELECTRIC SERVICE 293
six miles west of Mobile and the voltage of this line is only
2,300, it was feared that the mill would not be able to operate
more than one-half of the machines. The total connected horse-
power was 255. In order to raise the voltage nine 10-kw. trans-
formers which had their secondaries connected for 110 volts
were connected in series with the nine 20-kw. transformers in
such a way as to give a three-phase current having a voltage of
330 at no-load.
It was found that when the 75-hp. motor was started the
voltage dropped from 330 to 120. With this emergency bank
of transformers 210 hp. could be operated, and the mill was in
operation in the specified time.
It was decided as soon the contract was signed to extend
the 11,000-volt line from the power house directly to the mill.
The substation erected is very different from that at Whistler,
for the entire installation is exposed to the weather. The sub-
station consists of a wood platform eight feet above the ground,
forty-one feet long and seven and one-half feet wide, supported
on four thirty-five-foot poles and five posts. Between the first
and second poles, which are spaced eighteen feet apart, three
General Electric outdoor type multiplex lightning arresters are
hung on six-inch by eight-inch timbers. The second pole car-
ries the choke coils and series instrument transformers. The
third pole has the pole oil switch and shunt instrument trans-
formers, and between the third and fourth poles the main trans-
formers are located. These three transformers are rated at 75
kva., oil-cooled, and have taps on the high-tension side for
11,000, 10,800 and 10,600 volts. There are two sets of discon-
necting switches, one to cut out the lightning arresters and the
other to cut out all apparatus. The pole oil switch is so arranged
that it can be operated from the ground by pulling a sash cord.
The secondaries of the transformers pass down through the
floor and tap to two 500,000-circ. mil cables per phase. The
substation is about one hundred and twenty feet from the mill
and is located half way between the mill and a storage house,
thus being as far as possible from all building in case of fire.
THE ARMOUR ENGINEER
The Semi-Annual Technical Publication of the Student Body of
ARMOUR INSTITUTE OF TECHNOLOGY.
VOL. IV CHICAGO, MAY. 1912 XO. 2
Publishing Staff for the year 1912:
L. H. Roller, Editor.
M. A. Feiser, Business Manager. F. T. Bangs, Asst. Editor.
C. R. Leibrandt, Asso. Bus. Mgr. R. W. Ermeling, Asst. Bus. Mgr.
Board of Associate Editors:
H. M. Raymond, Dean of the Engineering Studies.
L. C. MoNiN, Dean of the Cultural Studies.
E. H. Freeman, Professor of Electrical Engineering.
G. F. Gebhardt, Professor of Mechanical Engineering.
H. McCormack, Professor of Chemical Engineering.
A. E. Phillips, Professor of Civil Engineering.
W. F. Shattuck,, Professor of Architecture.
F. Taylor, Professor of Fire Protection Engineering.
Published twice each year, in January and in May.
Publication office: Thirty-third St. and Armour Ave., Chicago, 111.
TERMS OF SUBSCRIPTION.
The Armour Engineer, two issues, postage prepaid $1.00 per aniuun
The technical press is invited to reproduce articles, or por-
tions of same, provided proper credit is given.
Every engineer, whether one in truth or an undergraduate
in the embryo stage, undoubtedly has an opinion as to the proper
course of training a graduate should pursue after receiving his
diploma. The successful practicing engineer has
The Young the advantage of being able to base his opinion
Engineer's on his own beginnings, while the graduate must
Training. look to the successes of others to see if he can
duplicate even one of them. The first step the
graduate takes is, in most cases, the most important one of his
business career, for it will generally spell success or failure for
May, 1912] EDITORIALS 295
him. For this reason there are considerations to which he must
give attention. To accept any position, with thought of only
the salary connected therewith, and no thought of what the fu-
ture will bring, is like striking in the dark ; the same results
may be expected.
For the young engineer to succeed practical training is nec-
essary. And this must include two things, either actual work-
ing experience "on the job," or work in an engineer's office. In
actual construction work he will acquire knowledge of engi-
neering operations, and of the time required for such opera-
tions, as well as quality and disposition of materials most suit-
able to meet various circumstances. While it is true that he can
not, during a short period, acquire anything approaching a great
knowledge of this side of professional work, still if use is m.ade
of fair opportunity he will be able to lay a sound foundation
for future experiences in this line. And to gain experience
which will best serve him in his training he must look at engi-
neering from a different angle — one which he will come to
realize the importance of in his construction work — that of
design.
Engineering as a profession is based upon the matter of
design. While it is true that large projects may be successfully
carried out by blindly following methods worn with usage, the
engineer, in carrying out the .same projects, would make them
the realization of well-laid and carefully-regulated design. While
results are more earnestly sought than projects, still the impor-
tance of designing under practical conditions cannot be over-
looked, and the young engineer should spend a part of his time
in an engineer's office as well as in actual construction work.
However good a man's training and experience in construction
or in shops may be, he cannot be an engineer without experience
in an engineer's office. He may make a good mechanic or con-
tractor, but will lack, perhaps sadly, experience with drafting
and design, specifications, costs and estimating.
On the other hand, the mere fact of being a member of the
staff in an engineer's office will not train a man to be an engi-
neer. He must see and deal with as many aspects and varieties
of work as possible, and must, above all, have training in or
296 THE ARMOUR ENGINEER [Vol. 4, No. 2
upon construction, or otherwise he will become a clerk, drafts-
man or commercial man, and not an engineer.
The possibility of becoming mired in a special, but rather
ordinary, line of work must not be overlooked. A case is called
to mind of a young civil engineer, who spent his first summer
in the field. Realizing the importance of design, he entered the
drafting department, to gain added experience. He has re-
mained there ten years, rising to the position of assistant chief
draftsman, but with small hope of advancement. To go back
to the field he would be compelled to start at the bottom, and
the thought of reduced salary keeps him where he is. He was
unfortunate in making his training all of one variety.
This thought leads up to the subject of specialization for
which it can be said that many of our best engineers are spe-
cialists in one line of work. In the majority of cases, however,
specialization has come after a varied training, not before. The
specialist has the training of the multitude before he is placed
upon his pedestal. For the young engineer specialization means
narrowness, and the graduate should make conscious effort to
secure a varied training to fit him for his profession.
F. T. B.
There are two great movements in the development of any
science. One is the rapidly increasing mass of isolated facts and
deductions ; this accumulation is so great and is increasing so
rapidly that no one person attempts to keep
The up with the developments in all their sig-
Conservatism nificances. The chemist, for example, be-
ef Matter comes either an inorganic or an organic
chemist. And then he limits himself to
a still smaller field, until he usually loses sight of the develop-
ments in the other fields of his own science.
The other movement is toward simplification and generahza-
tion. The mind gropes about for some simple generalization
which will correlate isolated facts. These generalizations have
proven of enormous value. One need only consider the influence
of the three great generalizations of chemistry, the atomic theory.
May, 1912] EDITORIALS
297
the periodic law, and the phase rule, upon that science to ilhis-
trate the importance of this movement.
Yet these generahzations, fruitful as they are, do not mean
much outside of the particular field of chemistry and that newer
development of the science, physical chemistry. On the other
hand the sciences are interrelated, and applications of one or
another are of every-day occurrence. Thus chemistry runs into
physics, engineering, geology, physiology, medicine, botany
zoology, and all the other sciences. Any generalization that ex-
tends throughout the whole field of science ought to be empha-
sized. We have, for example, the great simplifying generaliza-
tion, the importance of whose influence cannot be over-estimated.
Bancroft, to whom these ideas are largely due, has called
attention {Science, 1911. page 159) to another great generaliza-
tion, a quantitative one, it is true, but in its way just as signifi-
cant. To the chemist it is the theorem of Le Chatelier, to the
physicist, the principle of least action or the theory of De Mau-
pertuis, to the biologist the law of survival of the fittest to the
economist, the law of supply and demand. In mechanics 'we see
It again in the equality of action and reaction ; in electricity in
the phenomenon of induction. The broadest statement of 'the
generalization is that a system tends to change so as to minimize
an external disturbance.
"Nature does not make a jump." At the critical point a
vapor merges into Hquid without a break. Solid solutions and
liquid crystals have bridged the gap between the liquid and solid
states. Ostwald says that "if a system in equilibrium is subjected
to a constraint by which the equilibrium is shifted, a reaction
takes place which opposes the constant, i.e., one by which its
efi^ect is partially annulled." Van't Hoff showed that when a
system is in equilibrium, and its temperature is raised, any re-
action which takes place reduces the temperature, or at all events,
absorbs heat, and vice versa. He also showed that when the
pressure is increased any reaction which takes place tends to
reduce the volume, which under ordinary circumstances would
increase the pressure.
If we heat a liquid we convert a portion of it into vapor,
an operation which absorbs heat. If we heat a saturated solu-
298 THE ARMOUR ENGINEER [Vol. 4, No. 2
tion, the solubility increases if the solid dissolves with an ab-
sorption of heat. If we increase the pressure on a dissociating
compound, or if we increase the concentration of the dissocia-
tion products, we get a decrease in pressure and a decrease in
the amount of dissociation product. If we pass an electric cur-
rent through a solution we tend to get a counter electromotive-
force which cuts down the electrical stress.
If we have suspended particles in a liquid a difference of
potential causes them to move in the direction which reduces the
electrical stress. Since all substances absorb light of some wave
length to a greater or less extent, all substances are light sensi-
tive to some rays, and tend to change in such a way as to elimi-
nate the strain caused by the light. Thus with some silver salts
we get a visible decomposition. With chromium salts we get
no measurable change unless some reducing agent is present.
With some substances we get fluorescence or phosphorescence,
but all ordinarily without apparent change. With a copper sul-
phate solution there is apparently no effect due to light. Yet
all these substances are really light sensitive and they all tend
to change in the same way, namely to eliminate the substance
which absorbs the light.
In addition to these we have the law of maximum entropy,
which states that energy always tends to reduce itself to its low-
est form, namely, heat; and a portion of it may be expended in
bringing about a change which accelerates the degradation of
energy. Then there is Bischofif's "Dynamic Hypothesis," which
states that atoms always endeavor to take up positions which
give the most scope for vibration. Here the electrons associated
with the atoms are thus enabled to waste the largest amount of
energy in the form of radiant heat.
These laws and the general principle of conservatism apply
to all systems and changes of the condition of equilibrium, whether
physical or chemical, to evaporation and fusion, to solution and
chemical action. Also the system in changing does not pass from
one state directly to the most stable state suitable for the new
conditions. It passes from the most unstable state to the more
stable and gradually to the most stable.
May, 1912] EDITORIALS 299
We have seen the general applicability of the principle of
least action in the physical sciences. Bancroft did a valuable
service in showing the universal applicability of the idea, which
serves not only to bind facts in the various sciences together,
but also, and undoubtedly more important, to shed a clearer light
on the connection existing between phenomena in the different
sciences previously regarded as unrelated.
B. B. Freud.
300 THE ARMOUR ENGINEER [Vol. 4, No. 2
ARMOUR INSTITUTE OF TECHNOLOGY BRANCH
OF THE AMERICAN INSTITUTE OF
ELECTRICAL ENGINEERS.
Since the last issue of The Armour Engineer this society
has held several very interesting and instructive meetings. The
society has been very fortunate this semester in securing promi-
nent men in the electrical profession to address the meetings.
The first meeting for this semester was held February 14th,
at the Boston Oyster House. Dinner was served at 7 p. m.,
after which the meeting was addressed by Mr. P. G. Downton,
of the class of 1909, on "Engineering Application of Storage
Batteries." The talk was intended to give those present a gen-
eral idea of the correct application of storage batteries.
The next meeting was held in Chapin Hall, March 13th.
Mr. V. Pagliarulo, of the class of 1912, addressed the society
on "Electrical Equipment of the West Side Metropolitan Ele-
vated Railway." The speaker gave a general description of the
entire system and explained in detail many points of particular
interest.
The second March meeting was held in Chapin Hall, March
27th. Mr. S. H. Gushing, Statistician of the Public Service
Gompany of Northern Illinois, addressed the society on "The
Organization of a Public Service Corporation." Mr. Gushing's
knowledge of the organization of one of the largest corporations
in this country enabled him to show very clearly the need of
organization and to explain in detail the organization of a large
company.
On the evening of April 10th the annual banquet and elec-
tion of officers was held at Kuntz-Remmler's. The attendance
was especially good and everybody had a very pleasant time.
On Wednesday evening, April 24th, Mr. T. S. Stevens,
Signal Engineer for the Atchison, Topeka & Santa Fe Railroad,
addressed the society on "Signal Engineering." He illustrated
with drawings the development of signaling and explained the
purpose of signals. During the discussion Mr. J. E. Saunders,
of the class of 1907, Assistant Signal Engineer for the Atchison,
Topeka & Santa Fe Railroad, discussed briefly the use of alter-
nating current in signal work.
The past year has been a very busy one for this Branch.
With but three exceptions, two meetings have been held each
month. The topics discussed at these meetings have been of
general interest and every Senior and Junior feels that he has
been benefited.
— F. A. Graham.
May, 1912] ENGINEERING SOCIETIES 301
ARMOUR INSTITUTE OF TECHNOLOGY BRANCH
OF THE AMERICAN SOCIETY OF
MECHANICAL ENGINEERS.
The Armour Institute Student Branch of the American
Society of Mechanical Engineers has held six technical meet-
mgs this year, besides a smoker and a banquet, and expects to
close a successful season with a banquet at the Boston Oyster
House on May 15th. The meetings have been well attended,
and the active interest shown by Junior Class members gives
promise of success for next year.
On January 31, 1912, the Senior Class members of the Ar-
mour Branch were addressed by Mr. Calvin W. Rice, Secretary
of the A. S. M. E. He gave a most interesting talk, sketching
the many advantages to be derived from membership in the
A. S. M. E., and explaining how even student members might
get information upon any engineering subject by corresponding
with the librarian of the society's library in New York, the
largest technical library on the continent. Mr. Rice gave some
sound advice for the engineering school graduate and gave some
interesting illustrations from his own personal experience.
The first meeting of the second semester was held on Feb-
ruary 7, 1912. Mr. K. M. Boblett, '09, engineer for the Kinsey
Manufacturing Company, of Toledo, Ohio, spoke on "Automo-
bile Radiators." Mr. Boblett had a number of samples and
models of various types of radiators, which were minutely de-
scribed in the course of the lecture. The radiating effect of
each radiator, as well as the details of construction with refer-
ence to efficiency and economy of manufacture, were fully dwelt
upon. The subject was interesting, and the talk was enjoyed by
all present.
On March 6, 1912, Mr. Sydney V. James, '07, Consulting
Engineer for the Aero Club of Illinois, pres-ented a paper on
"Scientific Aeroplane Model Testing." The development of the
aeroplane model for experimental purposes was outlined, with
examples showing what close approximations have been made
to actual flying conditions in the past. It is believed that this
method will find more extensive use in the future, thus de-
creasing greatly the expense of aeronautic investigation. This
meeting was attended by about fifty persons.
On April 9, 1912. Mr. J. C. Miller, M. E.. presented a paper
on "Oil Engines." This meeting was held in Science Hall in
conjunction with the Chicago Branch of the Institute of Oper-
ating Engineers. Mr. Miller has given an unusual amount of
time and study to the oil engine, and gave a very interesting
talk, illustrated with lantern slides. The comparative cost of
various fuels was taken up, and it was shown that the lower
302 THE ARMOUR ENGINEER [Vol. 4, No. 2
grade fuel oil must .inevitably be drawn upon to a greater and
greater extent in the future. The latest developments in the
oil engine field were taken up with respect to both American
and European practice.
P. L. Keachie.
CIVIL ENGINEERING SOCIETY.
On the evening of January 16, 1912, Mr. N. W. Cloud,
Editor of the Signal Engineer, gave an illustrated lecture on
the "Relation of Signalling to the Civil Engineer." Mr. Cloud
clearly described signalling from the use of the semaphore to
the complicated interlocking systems and automatic block sys-
tems.
Col. Holp, of the Hollister Land Company, gave an illus-
trated commercial lecture on "California," in Science Hall, on
February 6, 1912. The important part of irrigation in the culti-
vation of the highlands in California was set forth. The Colonel
with his good slides and funny stories made everybody enthu-
siastic with the "California fever."
On March 5, 1912, Mr. Klein, '06, member of the firm of
Lieberman & Klein, Civil Engineers, addressed the society on
the "Practical Design of a Concrete Building." Many practical
points were noted by Mr. Klein which were of aid to the sen-
iors in their design of a "Concrete Warehouse."
Mr. W. Leininger, '06, Assistant Superintendent of Streets
in Chicago, on the evening of March 19, 1912, addressed the
society on "Street Repairs in Chicago." Mr. Leininger told of
the work carried out by the Bureau of Streets each year and the
amount of monev appropriated for the same. An interesting dis-
cussion followed on the use of oils on macadam roads for dust
prevention.
On April 2, 1912, Mr. Myron Reynolds, '06, gave an informal
talk on the "Practical and Theoretical Placing of Concrete."
He first took up the specifications required for the aggregate,
cement and sand necessary for a good mixture. He then dis-
cussed the methods of mixing and placing the same.
On Tuesday evening, April 16, 1912, Mr. J. C. Penn, '05,
gave an illustrated lecture on the "Substructure of the North
Avenue Bridge." This was another case of the speaker having
been "on the job,"' and the talk proved to be one of the best of
the year.
At the last meeting of the year, on May 7, 1912, Mr. Sid-
ney James, '07, Engineer of the Aero Club of Illinois, gave an
illustrated lecture on "Some Recent Results of Experimental In-
vestigation in Aeronautics." Mr. James described the labora-
May, 1912] ENGINEERING SOCIETIES 303
tory at the University of Paris and the Eiffel Laboratory for
the testing of aeroplane models. The disturbances of the air
caused by various shaped bodies passing through was noted and
their eft'ect on the design of aeroplanes. The most scientific and
safe way to aid in the development of aeronautics is to study a
small model in the laboratory, then after correcting all its faults
to build a large reproduction of the same.
Outside of the regular meetings of the society, on February
20, 1912, a smoker was held in connection with the election of
officers for the year 1912-13. On Friday evening, April 12, the
society held its' Annual Banquet at the Great Northern Hotel.
The words of advice from the Faculty were appreciated by all of
the fifty men present.
In conclusion we would say that the society has completed
the most successful and prosperous year in its existence, and
with the officers for next year we know it will attain a still
higher success. The Alumni are urged to get out as often as
possible and help the student members in their good work.
— C. W. Collins.
THE SENIOR CHEMICAL SOCIETY.
The monthly dinners of the Senior Chemical Society during
the second semester were just as popular with the members as
they were when the monthly banquet plan was first introduced
and their success was well shown by the enthusiasm of all of
the members and their regular attendance at the meetings.
On Jan. 17th, 1912, the Society held the first alumni dinner
of the school year at the Sherman House. About forty men
were present and short informal talks were given by all of the
alumni and the Profs. The chief speaker of the evening was
Prof. McCormack. A very good program was given by members
of the Glee and Mandolin Clubs, and many of the Armour songs
were sung, all joining in the chorus. After one of the most suc-
cessful dinners of the year, the meeting adjourned.
The sixth banquet of the Society was held Wednesday, Febru-
ary 14th, at Kuntz-Remmler's restaurant. The speaker of the
evening was Mr. VVm. Hoskins. the head of the firm of Mariner
& Hoskins, Consulting Chemists. His subject was "Factors for
Success as a Consulting Chemist." Mr. Hoskins gave a fine talk
on the conditions one might expect in starting life as a chemist
and also upon many legal questions which he has solved in his
capacity of consulting engineer. Twenty-five of the faculty and
members were present.
The seventh dinner of the Society was held at the Kaiserhof
Hotel on March 6. Mr. Loewenstein, Chief Chemist for Morris
304 THE ARMOUR ENGINEER [Vol. 4, No. 2
& Co., and Chairman of the Chicago Section of the American
Chemical Society, gave a snappy Httle talk on "The Chemical
Engineer in the Packing Industry."- The prospects of the chemi-
cal engineer and the value of his services were the chief points.
IVIany interesting facts were brought out by a discussion, which
at times assumed the character of a cross-examination. About
thirty-five members of the department attended this meeting.
The final dinner of the year was held in the Pine Room of
the Stratford Hotel on May 10. The speakers were prominent
alumni of the department, the faculty and the graduating under-
classmen. Several excellent speeches were made, among the best
being those of Professor McCormack, Mr. B. Hoffman, Mr.
Pulsifer and Mr. Tibbals. After a siege of stories, the meeting
was closed. About forty men attended this banquet, which
marked a fitting end to an excellent year for the Society.
S. Kahn.
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11